CoastSat Methodology: Experimental LLM Publication

Workflow Overview

Instructions

You are writing the introductory overview for the CoastSat workflow. Stay at a high level: explain the scientific objectives, data sources, and analysis themes without repeating the per-step prose that appears later in the document.

Metadata is provided as JSON with these keys:

  • site (site identifier used in the example)

  • total_steps

  • steps (ordered list of step dicts) where each step includes:

    • name, position, language, code_repository_markdown

    • inputs_summary, outputs_summary

    • inputs and outputs (each is a shortened list of parameter briefs with name, format, description, source_link, total_files, transient_note)

Return two short paragraphs (max ~160 words total). Requirements:

  • Mention the overall analytical aim and how satellite shoreline data is processed.

  • Highlight how the workflow progresses (input acquisition → processing → analysis → publication) referencing step names when useful.

  • Acknowledge the open-source nature by citing one or more code_repository_markdown links.

  • Do not invent steps or tools beyond the supplied list.

Workflow Metadata

Shared Context

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

CoastSat aims to monitor and model shoreline change by extracting shoreline positions from satellite imagery, correcting for tidal influence, estimating local topographic slope, and quantifying long-term trends. By transforming raw spatial observations into tidally adjusted, smoothed transect series, it enables consistent assessment of erosion, accretion, and stability across coastal sites.

The workflow proceeds through logical stages: data acquisition of shoreline polygons, transect geometries, and time-series measurements → correction and smoothing of shoreline positions (via tidal_correction.ipynb) → local slope estimation (via slope_estimation.ipynb) → statistical trend modeling (via linear_models.ipynb) → final summary output and formatting (via make_xlsx.py). All notebooks and scripts are open source in the UoA-eResearch CoastSat repository—see for example the tidal_correction.ipynb and make_xlsx.py code.

Workflow Diagram

Instructions

You are producing a diagrammatic summary of the CoastSat workflow for the introduction section. Use only the steps provided in the metadata.

Metadata keys:

  • step_order (list of step names in execution order)

  • steps (list of step dicts) with: name, position, code_repository_markdown, inputs, outputs

Return Markdown containing:

  1. A fenced code block (begin with and end with ) containing a flowchart TD diagram that links steps in order, annotating each node with (a) the step name and (b) a short phrase describing its primary role. When a step has a code_repository_markdown, include the link in brackets within the node label.

  2. A single-sentence caption beneath the diagram noting the kinds of artefacts produced (refer to outputs where available).

Guidelines:

  • Do not add steps or edges that are not reflected in step_order.

  • Keep node labels under 12 words and avoid backticks or braces inside labels.

  • Mermaid must be syntactically valid so the renderer can display the diagram.

Workflow Metadata

Shared Context

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

This workflow generates artefacts including tidally corrected time‐series CSVs, slope‐enhanced transect geometries, modelled shoreline change metrics, and Excel workbooks aggregating transect and site data.

Primary Outcomes

Instructions

You are writing a short overview of the primary outputs generated by the CoastSat workflow. Keep the focus on the final artefacts and what they enable for analysis.

Metadata keys:

  • site (site identifier in focus)

  • outcomes (list of output dicts) each containing:

    • step_name, parameter, name, format, description

    • sample_link (Markdown link when a public example exists)

    • total_files

    • transient_note (optional explanatory string)

Return a Markdown bullet list (max 5 bullets). Each bullet should:

  • Name the artefact (name or parameter) and state which step produces it (step_name).

  • Explain what the artefact represents using description or format.

  • Reference sample_link when available (these links already correspond to the provided site); otherwise state that the files are internal only.

  • Mention total_files to give a sense of scale.

  • Include transient_note verbatim when provided.

Conclude with a single sentence describing how these artefacts support downstream shoreline analysis.

Workflow Outcomes

Shared Context

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

Here is a refined overview of the primary artefacts generated by the CoastSat workflow for site nzd0001, focusing on the final outputs and what they enable in downstream analysis:

  • tides.csv (#fp-tides-5, produced by tidal_correction.ipynb): A table of observed and/or modelled tidal heights over time used to correct raw satellite-derived shoreline positions for tidal biases. Total files: ~580. Sample links are internal only for this site.

  • transect_time_series_tidally_corrected.csv (#fp-transecttimeseriestidallycorrected-2, produced by tidal_correction.ipynb): A time series of shoreline positions along each transect after applying tidal correction; enables detection of temporal changes in shoreline location. Total files: ~580.

  • transects_extended.geojson (#fp-transectsextended-3, produced by linear_models.ipynb): GeoJSON geometries of transects extended seaward/landward to enhance spatial coverage and consistency; supports spatial modelling of change. Total files: 1.

  • transects.xlsx (#fp-transects_xlsx-1, produced by make_xlsx.py): Spreadsheet listing each transect’s baseline geometry and metadata (e.g. IDs, orientation); useful for manual inspection, selection or mapping. Total files: 1.

  • site transect summary workbook (#fp-transect_site_xlsx-1, produced by make_xlsx.py): Excel workbook aggregating all transect time‐series and summary statistics (e.g. rates of change) for site nzd0001; supports reporting and comparative analyses. Total files: ~578.

Together, these artefacts allow for accurate, repeatable quantification of shoreline position changes over time (via tidal correction and time series), spatial modelling across transects, and accessible summaries for interpretation or management decisions.

Instructions

You are generating a concise title for a CoastSat computational workflow step.

Metadata is provided with fields such as name, position, language, code_repository, code_repository_markdown, and step_number (1-indexed order in the workflow).

Return the title as a level 2 Markdown heading in the format ## Step {step_number}: {short title}, embedding the code repository URL when available.

Step Metadata

Objective

Instructions

You are summarising the purpose of a single step in the CoastSat shoreline workflow.

Step metadata is provided as JSON with these keys:

  • identity (name, position, language, code_repository, and code_repository_markdown when a URL is supplied)

  • data_flows (input/output summaries, counts, examples, and links)

  • linked_artefacts (aggregated linked input/output details)

  • notebook.summary

  • context_lines (additional factual notes)

Return one sentence (max 30 words) that describes the step’s objective. Requirements:

  • Mention the main operation or outcome (e.g. “tidal correction”, “linear modelling”).

  • If the code repository URL is provided, include a Markdown link on the step name or main action.

  • Use factual, neutral language; no speculative claims.

  • Output only the sentence (no heading or bullet).

Operations

Instructions

You are outlining the major operations carried out in a CoastSat workflow step.

Step metadata is provided as JSON with these keys:

  • identity (name, position, language, code_repository, code_repository_markdown when present)

  • data_flows (input/output summaries, counts, examples, links)

  • linked_artefacts (linked input/output aggregates)

  • context_lines (additional factual notes)

  • notebook (overall summary plus cells, each with name, position, and full content when available)

Return a Markdown bullet list with two to four items. Requirements:

  • Each bullet should start with a verb phrase (e.g., “Download NIWA tide observations”).

  • Each bullet should be in order of execution within the notebook.

  • Reference notable inputs/outputs or notebook cells when they are provided; use Markdown links as supplied.

  • Keep each bullet under 20 words.

  • If a notebook cell would help describe the operation, include it by following the instructions below:

If a bulletpoint references a notebook cell, reference the cell name after the fullstop of the bulletpoint. For example: A bulletpoint point about something. Reference: [Code cell 3]. Next, include the cell content using the following syntax:

[CELL NAME HERE]

Note the use of > which will be rendered into a collapsible callout. Make sure there are no syntax errors in the formatting of the cell content -- note the "-" directly after [!note]. This must be present. Only reference a maxium of one notebook cell per bulletpoint.

Step Inputs

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

Step Outputs

Instructions

You are documenting an output artefact produced by a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 35 words) describing what this output provides and how downstream work can use it. Requirements:

  • Begin with the artefact name (name) and explain the product using description or format.

  • Cite the public source_link when available.

  • Reference the contributing notebook cells via cell_refs (or state cell_refs_note when none are documented).

  • If linked_examples contains a link, mention the first example inline and reference total_linked_files for scale.

  • Weave in linked_files_note and transient_note (when present) without speculation.

  • Use only supplied metadata; do not invent results or performance claims.

Instructions

You are generating a concise title for a CoastSat computational workflow step.

Metadata is provided with fields such as name, position, language, code_repository, code_repository_markdown, and step_number (1-indexed order in the workflow).

Return the title as a level 2 Markdown heading in the format ## Step {step_number}: {short title}, embedding the code repository URL when available.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

Step 1: Tidal Correction – applying water level adjustments using tide records via tidal_correction.ipynb

Step Metadata

Objective

Instructions

You are summarising the purpose of a single step in the CoastSat shoreline workflow.

Step metadata is provided as JSON with these keys:

  • identity (name, position, language, code_repository, and code_repository_markdown when a URL is supplied)

  • data_flows (input/output summaries, counts, examples, and links)

  • linked_artefacts (aggregated linked input/output details)

  • notebook.summary

  • context_lines (additional factual notes)

Return one sentence (max 30 words) that describes the step’s objective. Requirements:

  • Mention the main operation or outcome (e.g. “tidal correction”, “linear modelling”).

  • If the code repository URL is provided, include a Markdown link on the step name or main action.

  • Use factual, neutral language; no speculative claims.

  • Output only the sentence (no heading or bullet).

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

The purpose of the tidal_correction.ipynb step is to adjust transect time-series shoreline positions using tide data (i.e. applying tidal correction).

Operations

Instructions

You are outlining the major operations carried out in a CoastSat workflow step.

Step metadata is provided as JSON with these keys:

  • identity (name, position, language, code_repository, code_repository_markdown when present)

  • data_flows (input/output summaries, counts, examples, links)

  • linked_artefacts (linked input/output aggregates)

  • context_lines (additional factual notes)

  • notebook (overall summary plus cells, each with name, position, and full content when available)

Return a Markdown bullet list with two to four items. Requirements:

  • Each bullet should start with a verb phrase (e.g., “Download NIWA tide observations”).

  • Each bullet should be in order of execution within the notebook.

  • Reference notable inputs/outputs or notebook cells when they are provided; use Markdown links as supplied.

  • Keep each bullet under 20 words.

  • If a notebook cell would help describe the operation, include it by following the instructions below:

If a bulletpoint references a notebook cell, reference the cell name after the fullstop of the bulletpoint. For example: A bulletpoint point about something. Reference: [Code cell 3]. Next, include the cell content using the following syntax:

[CELL NAME HERE]

Note the use of > which will be rendered into a collapsible callout. Make sure there are no syntax errors in the formatting of the cell content -- note the "-" directly after [!note]. This must be present. Only reference a maxium of one notebook cell per bulletpoint.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

  • Read and index shoreline polygons from polygons.geojson, selecting those with “nzd” IDs. [Code cell 2]

Code cell 2

  • Load transect time series per site, extract rounded timestamps and compute site centroid. [Code cell 3]

Code cell 3

  • Request tide heights from NIWA API at each timestamp for sites missing tide data; save as tides.csv. [Code cell 6]

Code cell 6

Step Inputs

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

#fp-polygons-1

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-polygons-1 is a GeoJSON file containing polygon geometries (“polygons.geojson”) used by the tidal_correction.ipynb step to spatially identify shoreline sites for correcting transect time series, as referenced in the notebook cells: not documented; 1 linked file referenced in the crate; representative example: polygons.geojson.

#fp-transecttimeseries-1

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-transecttimeseries-1 is a CSV input artefact containing shoreline-transect time-series data used in tidal_correction.ipynb (notebook cells 1-3) to apply tidal adjustments; 3121 linked files referenced in the crate; representative example: transect_time_series.csv

#fp-transectsextended-1

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-transectsextended-1 is a GeoJSON file containing the extended transect geometries used in the tidal correction step of tidal_correction.ipynb (see cells: Notebook cells: not documented); 1 linked file referenced in the crate; representative example: transects_extended.geojson.

#fp-tides-1

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-tides-1 is a tides data input artefact containing CSV-formatted tidal height time series (580 linked files; representative example: tides.csv) used in the tidal_correction.ipynb step (notebook cells: not documented.).

Step Outputs

Instructions

You are documenting an output artefact produced by a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 35 words) describing what this output provides and how downstream work can use it. Requirements:

  • Begin with the artefact name (name) and explain the product using description or format.

  • Cite the public source_link when available.

  • Reference the contributing notebook cells via cell_refs (or state cell_refs_note when none are documented).

  • If linked_examples contains a link, mention the first example inline and reference total_linked_files for scale.

  • Weave in linked_files_note and transient_note (when present) without speculation.

  • Use only supplied metadata; do not invent results or performance claims.

#fp-tides-2

Instructions

You are documenting an output artefact produced by a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 35 words) describing what this output provides and how downstream work can use it. Requirements:

  • Begin with the artefact name (name) and explain the product using description or format.

  • Cite the public source_link when available.

  • Reference the contributing notebook cells via cell_refs (or state cell_refs_note when none are documented).

  • If linked_examples contains a link, mention the first example inline and reference total_linked_files for scale.

  • Weave in linked_files_note and transient_note (when present) without speculation.

  • Use only supplied metadata; do not invent results or performance claims.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-tides-2 provides tide‐height time‐series data (format: CSV or similar) used by subsequent steps (e.g. producing #fp-transecttimeseriestidallycorrected-1) to apply tidal corrections to shoreline transect measurements; it is marked transient and currently has no linked files recorded. (cells: Notebook cells: not documented; source: tidal_correction.ipynb)

#fp-tides-3

Instructions

You are documenting an output artefact produced by a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 35 words) describing what this output provides and how downstream work can use it. Requirements:

  • Begin with the artefact name (name) and explain the product using description or format.

  • Cite the public source_link when available.

  • Reference the contributing notebook cells via cell_refs (or state cell_refs_note when none are documented).

  • If linked_examples contains a link, mention the first example inline and reference total_linked_files for scale.

  • Weave in linked_files_note and transient_note (when present) without speculation.

  • Use only supplied metadata; do not invent results or performance claims.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-tides-3 is a transient output from tidal_correction.ipynb providing corrected tide data (in its format as per the workflow definition), usable by downstream steps needing tide heights; it is generated in the notebook’s initial cells and links no physical files.

#fp-transecttimeseriestidallycorrected-1

Instructions

You are documenting an output artefact produced by a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 35 words) describing what this output provides and how downstream work can use it. Requirements:

  • Begin with the artefact name (name) and explain the product using description or format.

  • Cite the public source_link when available.

  • Reference the contributing notebook cells via cell_refs (or state cell_refs_note when none are documented).

  • If linked_examples contains a link, mention the first example inline and reference total_linked_files for scale.

  • Weave in linked_files_note and transient_note (when present) without speculation.

  • Use only supplied metadata; do not invent results or performance claims.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-transecttimeseriestidallycorrected-1 provides the transect time-series data corrected for tidal variation as output from the tidal_correction.ipynb workflow step (coded in Python via tidal_correction.ipynb), enabling downstream analyses of shoreline change independent of tidal effects; cell_refs_note applies.

Instructions

You are generating a concise title for a CoastSat computational workflow step.

Metadata is provided with fields such as name, position, language, code_repository, code_repository_markdown, and step_number (1-indexed order in the workflow).

Return the title as a level 2 Markdown heading in the format ## Step {step_number}: {short title}, embedding the code repository URL when available.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

Step 2: Estimate Shoreline Slope (slope_estimation.ipynb)

Step Metadata

Objective

Instructions

You are summarising the purpose of a single step in the CoastSat shoreline workflow.

Step metadata is provided as JSON with these keys:

  • identity (name, position, language, code_repository, and code_repository_markdown when a URL is supplied)

  • data_flows (input/output summaries, counts, examples, and links)

  • linked_artefacts (aggregated linked input/output details)

  • notebook.summary

  • context_lines (additional factual notes)

Return one sentence (max 30 words) that describes the step’s objective. Requirements:

  • Mention the main operation or outcome (e.g. “tidal correction”, “linear modelling”).

  • If the code repository URL is provided, include a Markdown link on the step name or main action.

  • Use factual, neutral language; no speculative claims.

  • Output only the sentence (no heading or bullet).

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

The purpose of the slope_estimation.ipynb step is to compute missing beach slope values by analysing transects and time-series data, producing an updated extended transects file.

Operations

Instructions

You are outlining the major operations carried out in a CoastSat workflow step.

Step metadata is provided as JSON with these keys:

  • identity (name, position, language, code_repository, code_repository_markdown when present)

  • data_flows (input/output summaries, counts, examples, links)

  • linked_artefacts (linked input/output aggregates)

  • context_lines (additional factual notes)

  • notebook (overall summary plus cells, each with name, position, and full content when available)

Return a Markdown bullet list with two to four items. Requirements:

  • Each bullet should start with a verb phrase (e.g., “Download NIWA tide observations”).

  • Each bullet should be in order of execution within the notebook.

  • Reference notable inputs/outputs or notebook cells when they are provided; use Markdown links as supplied.

  • Keep each bullet under 20 words.

  • If a notebook cell would help describe the operation, include it by following the instructions below:

If a bulletpoint references a notebook cell, reference the cell name after the fullstop of the bulletpoint. For example: A bulletpoint point about something. Reference: [Code cell 3]. Next, include the cell content using the following syntax:

[CELL NAME HERE]

Note the use of > which will be rendered into a collapsible callout. Make sure there are no syntax errors in the formatting of the cell content -- note the "-" directly after [!note]. This must be present. Only reference a maxium of one notebook cell per bulletpoint.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

Here are the major operations in the slope_estimation.ipynb step, in order:

  • Load required modules and slope-estimation library. Reference: [Code cell 1].

Code cell 1

  • Read geospatial transects file and index by ID.

  • Select transects needing slope estimates: those with specific ID pattern, above an ID threshold, and missing slope data.

  • Loop through each selected transect: load its time-series + tide data, compute spectral slope estimates & confidence intervals, update transects, then save updated GeoJSON.

Step Inputs

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

#fp-transectsextended-1

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-transectsextended-1 is a GeoJSON file with extended transect geometries, used as input in the slope_estimation.ipynb workflow; see transects_extended.geojson; notebook cells: not documented; 1 linked file in the crate; representative example: transects_extended.geojson.

#fp-transecttimeseries-1

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-transecttimeseries-1 is a CSV‐format input artefact representing time‐series measurements of shoreline positions along transects; used by slope_estimation.ipynb (Python notebook) in cells listed below; 3121 linked files referenced in the crate; representative example: transect_time_series.csv.

#fp-tides-3

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-tides-3 represents the tide data input parameter (format assumed time-series or CSV) used by the slope_estimation.ipynb step; referenced in upstream cells (Notebook cells: not documented); No linked files recorded; the crate may not capture intermediate artefacts.

Step Outputs

Instructions

You are documenting an output artefact produced by a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 35 words) describing what this output provides and how downstream work can use it. Requirements:

  • Begin with the artefact name (name) and explain the product using description or format.

  • Cite the public source_link when available.

  • Reference the contributing notebook cells via cell_refs (or state cell_refs_note when none are documented).

  • If linked_examples contains a link, mention the first example inline and reference total_linked_files for scale.

  • Weave in linked_files_note and transient_note (when present) without speculation.

  • Use only supplied metadata; do not invent results or performance claims.

#fp-transectsextended-2

Instructions

You are documenting an output artefact produced by a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 35 words) describing what this output provides and how downstream work can use it. Requirements:

  • Begin with the artefact name (name) and explain the product using description or format.

  • Cite the public source_link when available.

  • Reference the contributing notebook cells via cell_refs (or state cell_refs_note when none are documented).

  • If linked_examples contains a link, mention the first example inline and reference total_linked_files for scale.

  • Weave in linked_files_note and transient_note (when present) without speculation.

  • Use only supplied metadata; do not invent results or performance claims.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-transectsextended-2 provides updated transect geometries in GeoJSON format with newly estimated beach slopes, built from transects_extended.geojson, transect time-series, and tide inputs via the slope_estimation.ipynb workflow step (code available at slope_estimation.ipynb); it is used downstream to enable slope-informed shoreline change analysis.

Instructions

You are generating a concise title for a CoastSat computational workflow step.

Metadata is provided with fields such as name, position, language, code_repository, code_repository_markdown, and step_number (1-indexed order in the workflow).

Return the title as a level 2 Markdown heading in the format ## Step {step_number}: {short title}, embedding the code repository URL when available.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

Step 3: Tidal Correction (tidal_correction.ipynb)

Step Metadata

Objective

Instructions

You are summarising the purpose of a single step in the CoastSat shoreline workflow.

Step metadata is provided as JSON with these keys:

  • identity (name, position, language, code_repository, and code_repository_markdown when a URL is supplied)

  • data_flows (input/output summaries, counts, examples, and links)

  • linked_artefacts (aggregated linked input/output details)

  • notebook.summary

  • context_lines (additional factual notes)

Return one sentence (max 30 words) that describes the step’s objective. Requirements:

  • Mention the main operation or outcome (e.g. “tidal correction”, “linear modelling”).

  • If the code repository URL is provided, include a Markdown link on the step name or main action.

  • Use factual, neutral language; no speculative claims.

  • Output only the sentence (no heading or bullet).

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

The tidal_correction.ipynb step performs a tidal correction by computing tides for each transect and adjusting the raw shoreline time series accordingly.

Operations

Instructions

You are outlining the major operations carried out in a CoastSat workflow step.

Step metadata is provided as JSON with these keys:

  • identity (name, position, language, code_repository, code_repository_markdown when present)

  • data_flows (input/output summaries, counts, examples, links)

  • linked_artefacts (linked input/output aggregates)

  • context_lines (additional factual notes)

  • notebook (overall summary plus cells, each with name, position, and full content when available)

Return a Markdown bullet list with two to four items. Requirements:

  • Each bullet should start with a verb phrase (e.g., “Download NIWA tide observations”).

  • Each bullet should be in order of execution within the notebook.

  • Reference notable inputs/outputs or notebook cells when they are provided; use Markdown links as supplied.

  • Keep each bullet under 20 words.

  • If a notebook cell would help describe the operation, include it by following the instructions below:

If a bulletpoint references a notebook cell, reference the cell name after the fullstop of the bulletpoint. For example: A bulletpoint point about something. Reference: [Code cell 3]. Next, include the cell content using the following syntax:

[CELL NAME HERE]

Note the use of > which will be rendered into a collapsible callout. Make sure there are no syntax errors in the formatting of the cell content -- note the "-" directly after [!note]. This must be present. Only reference a maxium of one notebook cell per bulletpoint.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

  • Read shoreline polygons, filter those with IDs starting with “nzd”, and index by ID. Reference: [Code cell 2]

    Code cell 2

  • For each transect time series per site, round timestamps to 10-minute intervals and find centroid of the corresponding polygon. Reference: [Code cell 3]

    Code cell 3

  • Pull tide data from NIWA API for each rounded datetime and site centroid; handle rate limits and retries. Reference: [Code cell 6]

    Code cell 6

  • Save daily tide values per site to tides.csv and output the tidally corrected transect time series.

Step Inputs

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

#fp-polygons-1

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-polygons-1 is a GeoJSON file providing coastline polygon geometries as spatial inputs for the tidal_correction.ipynb step (a Python notebook; step 2 in position) used in the code cells involved in reading and selecting polygons from polygons.geojson (see notebook cells 2-3). It has 1 linked file referenced in the crate; representative example: [polygons.geojson]citeturn0turn0

#fp-transecttimeseries-1

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-transecttimeseries-1 is a set of CSV files containing shoreline‐transect time‐series data used as input to tidal_correction.ipynb; example file: transect_time_series.csv; 3121 linked files referenced in the crate; notebook cells: not documented.

#fp-transectsextended-2

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-transectsextended-2 is an input artefact representing an extended transect time-series dataset (CSV-like timeseries with transect geometry and measurements), used by tidal_correction.ipynb (step 2), where it is consumed in cells 1–3; no linked files recorded (“No linked files referenced for this parameter; the crate may not capture intermediate artefacts.”).

#fp-tides-3

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-tides-3 is the named input parameter for tidal correction step representing a tides dataset (format: CSV) used within tidal_correction.ipynb (Python, workflow position 2), sourced from source_link, and lacks any linked example files; the crate records no linked files, and no notebook cells are documented (cell_refs_note).

Step Outputs

Instructions

You are documenting an output artefact produced by a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 35 words) describing what this output provides and how downstream work can use it. Requirements:

  • Begin with the artefact name (name) and explain the product using description or format.

  • Cite the public source_link when available.

  • Reference the contributing notebook cells via cell_refs (or state cell_refs_note when none are documented).

  • If linked_examples contains a link, mention the first example inline and reference total_linked_files for scale.

  • Weave in linked_files_note and transient_note (when present) without speculation.

  • Use only supplied metadata; do not invent results or performance claims.

#fp-tides-4

Instructions

You are documenting an output artefact produced by a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 35 words) describing what this output provides and how downstream work can use it. Requirements:

  • Begin with the artefact name (name) and explain the product using description or format.

  • Cite the public source_link when available.

  • Reference the contributing notebook cells via cell_refs (or state cell_refs_note when none are documented).

  • If linked_examples contains a link, mention the first example inline and reference total_linked_files for scale.

  • Weave in linked_files_note and transient_note (when present) without speculation.

  • Use only supplied metadata; do not invent results or performance claims.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

The output artefact #fp-tides-4 provides the CSV file tides.csv (format: comma-separated values) containing the calculated tidal correction values, which downstream steps (e.g. applying to shoreline position timeseries) can use to adjust transect elevation or water level records; this artefact is produced in step tidal_correction.ipynb (cells: cell_1 → cell_3), has no linked files in the crate (transient intermediate), and complements other outputs like transect_time_series_tidally_corrected.csv among the ~1160 files generated.

#fp-tides-5

Instructions

You are documenting an output artefact produced by a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 35 words) describing what this output provides and how downstream work can use it. Requirements:

  • Begin with the artefact name (name) and explain the product using description or format.

  • Cite the public source_link when available.

  • Reference the contributing notebook cells via cell_refs (or state cell_refs_note when none are documented).

  • If linked_examples contains a link, mention the first example inline and reference total_linked_files for scale.

  • Weave in linked_files_note and transient_note (when present) without speculation.

  • Use only supplied metadata; do not invent results or performance claims.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-tides-5 produces a set of CSV files (tides.csv etc.) containing calculated tidal offsets for each site, which downstream steps (e.g. shoreline correction) use to adjust raw transect elevations; source code in tidal_correction.ipynb underpins the process, linked examples include tides.csv among ~ 580 files; notebook cells: not documented.

#fp-transecttimeseriestidallycorrected-2

Instructions

You are documenting an output artefact produced by a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 35 words) describing what this output provides and how downstream work can use it. Requirements:

  • Begin with the artefact name (name) and explain the product using description or format.

  • Cite the public source_link when available.

  • Reference the contributing notebook cells via cell_refs (or state cell_refs_note when none are documented).

  • If linked_examples contains a link, mention the first example inline and reference total_linked_files for scale.

  • Weave in linked_files_note and transient_note (when present) without speculation.

  • Use only supplied metadata; do not invent results or performance claims.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-transecttimeseriestidallycorrected-2 provides transect time-series data (CSV format) with shoreline positions adjusted for tidal effects, produced in step tidal_correction.ipynb (workflow position 2, Python) using inputs like transect time series and polygons; downstream analyses can use these 580 corrected files (e.g. [transect_time_series_tidally_corrected.csv]) for modeling shoreline change over time (notebook cells not documented).

Instructions

You are generating a concise title for a CoastSat computational workflow step.

Metadata is provided with fields such as name, position, language, code_repository, code_repository_markdown, and step_number (1-indexed order in the workflow).

Return the title as a level 2 Markdown heading in the format ## Step {step_number}: {short title}, embedding the code repository URL when available.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

Step 4: Fit Linear Shoreline‐Change Models (linear_models.ipynb)

Step Metadata

Objective

Instructions

You are summarising the purpose of a single step in the CoastSat shoreline workflow.

Step metadata is provided as JSON with these keys:

  • identity (name, position, language, code_repository, and code_repository_markdown when a URL is supplied)

  • data_flows (input/output summaries, counts, examples, and links)

  • linked_artefacts (aggregated linked input/output details)

  • notebook.summary

  • context_lines (additional factual notes)

Return one sentence (max 30 words) that describes the step’s objective. Requirements:

  • Mention the main operation or outcome (e.g. “tidal correction”, “linear modelling”).

  • If the code repository URL is provided, include a Markdown link on the step name or main action.

  • Use factual, neutral language; no speculative claims.

  • Output only the sentence (no heading or bullet).

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

The notebook linear_models.ipynb fits linear regression models to tidally corrected time-series data along transects to compute shoreline change trends and outputs an extended transects GeoJSON.

Operations

Instructions

You are outlining the major operations carried out in a CoastSat workflow step.

Step metadata is provided as JSON with these keys:

  • identity (name, position, language, code_repository, code_repository_markdown when present)

  • data_flows (input/output summaries, counts, examples, links)

  • linked_artefacts (linked input/output aggregates)

  • context_lines (additional factual notes)

  • notebook (overall summary plus cells, each with name, position, and full content when available)

Return a Markdown bullet list with two to four items. Requirements:

  • Each bullet should start with a verb phrase (e.g., “Download NIWA tide observations”).

  • Each bullet should be in order of execution within the notebook.

  • Reference notable inputs/outputs or notebook cells when they are provided; use Markdown links as supplied.

  • Keep each bullet under 20 words.

  • If a notebook cell would help describe the operation, include it by following the instructions below:

If a bulletpoint references a notebook cell, reference the cell name after the fullstop of the bulletpoint. For example: A bulletpoint point about something. Reference: [Code cell 3]. Next, include the cell content using the following syntax:

[CELL NAME HERE]

Note the use of > which will be rendered into a collapsible callout. Make sure there are no syntax errors in the formatting of the cell content -- note the "-" directly after [!note]. This must be present. Only reference a maxium of one notebook cell per bulletpoint.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

Here are the major operations in the linear_models.ipynb step, in execution order:

  • Load required Python modules and prepare regression framework. Reference: [Code cell 1].

    [Code cell 1]

  • Read extended transect geometries (with beach slope) into GeoDataFrame, index by transect ID. Reference: [Code cell 2].

    [Code cell 2]

  • Gather file paths for VOS time series CSVs excluding certain patterns. Reference: [Code cell 3].

    [Code cell 3]

  • Combine file lists from multiple sources (tidally corrected and original transect time series) into unified sequence. Reference: [Code cell 6].

    [Code cell 6]

  • Perform linear regression per transect using the time-series inputs to extend transect geometries. Output saved as transects_extended.geojson.

Step Inputs

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

#fp-transectsextended-2

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-transectsextended-2 represents the GeoJSON file “transects_extended.geojson”, containing extended transect geometries used in notebook cells 1-3 for linear regression models in linear_models.ipynb; No linked files recorded, and this artefact may be transient.

#fp-timeseriestidallycorrected-1

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-timeseriestidallycorrected-1 is a CSV-format input artefact representing tide‐corrected time series data per transect (e.g. time_series_tidally_corrected.csv) used in linear_models.ipynb (cells 2–3) for fitting shoreline change trends; 3071 linked files referenced in the crate; representative example: time_series_tidally_corrected.csv.

#fp-transecttimeseriestidallycorrected-2

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-transecttimeseriestidallycorrected-2 is a CSV input artefact providing tidally corrected shoreline position time‐series for individual transects (format: CSV; 580 linked files in the crate; representative example transect_time_series_tidally_corrected.csv) used in notebook linear_models.ipynb (cells 1-3) for modeling linear trends.

#fp-transecttimeseries-1

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-transecttimeseries-1 represents a CSV-format time-series of transect shoreline positions (3121 files referenced; linked example transect_time_series.csv), used in linear_models.ipynb (step 3, Python) within code cells including “from glob … transect_time_series.csv” to fit linear regression models.

Step Outputs

Instructions

You are documenting an output artefact produced by a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 35 words) describing what this output provides and how downstream work can use it. Requirements:

  • Begin with the artefact name (name) and explain the product using description or format.

  • Cite the public source_link when available.

  • Reference the contributing notebook cells via cell_refs (or state cell_refs_note when none are documented).

  • If linked_examples contains a link, mention the first example inline and reference total_linked_files for scale.

  • Weave in linked_files_note and transient_note (when present) without speculation.

  • Use only supplied metadata; do not invent results or performance claims.

#fp-transectsextended-3

Instructions

You are documenting an output artefact produced by a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 35 words) describing what this output provides and how downstream work can use it. Requirements:

  • Begin with the artefact name (name) and explain the product using description or format.

  • Cite the public source_link when available.

  • Reference the contributing notebook cells via cell_refs (or state cell_refs_note when none are documented).

  • If linked_examples contains a link, mention the first example inline and reference total_linked_files for scale.

  • Weave in linked_files_note and transient_note (when present) without speculation.

  • Use only supplied metadata; do not invent results or performance claims.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

The output artefact #fp-transectsextended-3 is the file transects_extended.geojson, a GeoJSON representation of extended transect geometries derived from linear models in the linear_models.ipynb step (Position 3; Python), usable for downstream spatial analyses and change detection; see example [transects_extended.geojson] among the 1 linked file.

Instructions

You are generating a concise title for a CoastSat computational workflow step.

Metadata is provided with fields such as name, position, language, code_repository, code_repository_markdown, and step_number (1-indexed order in the workflow).

Return the title as a level 2 Markdown heading in the format ## Step {step_number}: {short title}, embedding the code repository URL when available.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

Step 5: Generate XLSX Summary with make_xlsx.py make_xlsx.py

Step Metadata

Objective

Instructions

You are summarising the purpose of a single step in the CoastSat shoreline workflow.

Step metadata is provided as JSON with these keys:

  • identity (name, position, language, code_repository, and code_repository_markdown when a URL is supplied)

  • data_flows (input/output summaries, counts, examples, and links)

  • linked_artefacts (aggregated linked input/output details)

  • notebook.summary

  • context_lines (additional factual notes)

Return one sentence (max 30 words) that describes the step’s objective. Requirements:

  • Mention the main operation or outcome (e.g. “tidal correction”, “linear modelling”).

  • If the code repository URL is provided, include a Markdown link on the step name or main action.

  • Use factual, neutral language; no speculative claims.

  • Output only the sentence (no heading or bullet).

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

The purpose of the step is to generate Excel‐formatted summaries by converting the GeoJSON transects and tidally corrected time‐series CSV inputs into unified .xlsx files (site‐level and overall) for further analysis.

Operations

Instructions

You are outlining the major operations carried out in a CoastSat workflow step.

Step metadata is provided as JSON with these keys:

  • identity (name, position, language, code_repository, code_repository_markdown when present)

  • data_flows (input/output summaries, counts, examples, links)

  • linked_artefacts (linked input/output aggregates)

  • context_lines (additional factual notes)

  • notebook (overall summary plus cells, each with name, position, and full content when available)

Return a Markdown bullet list with two to four items. Requirements:

  • Each bullet should start with a verb phrase (e.g., “Download NIWA tide observations”).

  • Each bullet should be in order of execution within the notebook.

  • Reference notable inputs/outputs or notebook cells when they are provided; use Markdown links as supplied.

  • Keep each bullet under 20 words.

  • If a notebook cell would help describe the operation, include it by following the instructions below:

If a bulletpoint references a notebook cell, reference the cell name after the fullstop of the bulletpoint. For example: A bulletpoint point about something. Reference: [Code cell 3]. Next, include the cell content using the following syntax:

[CELL NAME HERE]

Note the use of > which will be rendered into a collapsible callout. Make sure there are no syntax errors in the formatting of the cell content -- note the "-" directly after [!note]. This must be present. Only reference a maxium of one notebook cell per bulletpoint.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

  • Read extended transects GeoJSON from #fp-transectsextended-1 together with all tidally-corrected transect time-series CSVs from #fp-transecttimeseriestidallycorrected-2.

  • Compile combined transect data into a workbook saved as transects.xlsx in #fp-transects_xlsx-1.

  • For each site (e.g., nzd0058, nzd0483…), extract its site-specific time series and metadata.

  • Write each site into its own Excel file (e.g., nzd0058.xlsx, …) under parameter #fp-transect_site_xlsx-1.

Step Inputs

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

#fp-transectsextended-1

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-transectsextended-1 is the extended transect geometry GeoJSON file serving as one of the two inputs to the Python step make_xlsx.py (workflow position 4) to generate Excel outputs; representative example: transects_extended.geojson; 1 linked file referenced in the crate; Notebook cells: not documented.

#fp-transecttimeseriestidallycorrected-2

Instructions

You are documenting an input artefact for a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 30 words) describing the role of this input in the step. Requirements:

  • Begin with the artefact name (name) and describe what it represents using description or format.

  • Cite the public source_link when it is a Markdown link.

  • Mention the relevant notebook cells using cell_refs (or reuse cell_refs_note when empty).

  • If linked_examples contains a link, mention the first example inline.

  • Incorporate linked_files_note verbatim, and acknowledge transient_note if present.

  • Use only supplied information; do not speculate.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-transecttimeseriestidallycorrected-2 is a CSV file input showing tidally-corrected shoreline positions over time for multiple transects (580 linked files; representative example: transect_time_series_tidally_corrected.csv) used by the Python script make_xlsx.py to generate Excel outputs. Notebook cells: not documented.

Step Outputs

Instructions

You are documenting an output artefact produced by a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 35 words) describing what this output provides and how downstream work can use it. Requirements:

  • Begin with the artefact name (name) and explain the product using description or format.

  • Cite the public source_link when available.

  • Reference the contributing notebook cells via cell_refs (or state cell_refs_note when none are documented).

  • If linked_examples contains a link, mention the first example inline and reference total_linked_files for scale.

  • Weave in linked_files_note and transient_note (when present) without speculation.

  • Use only supplied metadata; do not invent results or performance claims.

#fp-transects_xlsx-1

Instructions

You are documenting an output artefact produced by a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 35 words) describing what this output provides and how downstream work can use it. Requirements:

  • Begin with the artefact name (name) and explain the product using description or format.

  • Cite the public source_link when available.

  • Reference the contributing notebook cells via cell_refs (or state cell_refs_note when none are documented).

  • If linked_examples contains a link, mention the first example inline and reference total_linked_files for scale.

  • Weave in linked_files_note and transient_note (when present) without speculation.

  • Use only supplied metadata; do not invent results or performance claims.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-transects_xlsx-1 produces transects.xlsx, a spreadsheet compiling combined transect-timeseries and tidally-corrected shoreline data into a structured Excel workbook using inputs from transects_extended.geojson and transect_time_series_tidally_corrected.csv; see make_xlsx.py for implementation, and refers to notebook cells: {cell_refs_note}.

#fp-transect_site_xlsx-1

Instructions

You are documenting an output artefact produced by a CoastSat workflow step.

Metadata is provided as JSON with these keys:

  • step (name, position, language, code_repository_markdown when available)

  • parameter, name, format, description

  • source_link (Markdown link when a public URL exists, otherwise plain name)

  • linked_examples (list of public Markdown links)

  • total_linked_files (integer count of linked files in the crate)

  • linked_files_note (plain-text description of linked-file availability)

  • transient_note (optional note when the artefact is a transient, non-materialised state)

  • cell_refs (list of notebook cell names), cell_refs_note (fallback sentence)

  • context_lines (supplementary bullet-style facts about the step)

Return one sentence (max 35 words) describing what this output provides and how downstream work can use it. Requirements:

  • Begin with the artefact name (name) and explain the product using description or format.

  • Cite the public source_link when available.

  • Reference the contributing notebook cells via cell_refs (or state cell_refs_note when none are documented).

  • If linked_examples contains a link, mention the first example inline and reference total_linked_files for scale.

  • Weave in linked_files_note and transient_note (when present) without speculation.

  • Use only supplied metadata; do not invent results or performance claims.

CoastSat Overview

CoastSat is a computational workflow for extracting and analysing shoreline positions from satellite imagery. It combines NIWA tide records, transect geometries, and Python notebooks to correct, smooth, and model shoreline change over time. Data and notebooks are versioned in GitHub (see the UoA-eResearch/CoastSat repository) and packaged as RO-Crates inside interface.crate. Use this context when describing individual workflow steps, artefacts, or intermediate data.

#fp-transect_site_xlsx-1 produces per-site Excel (.xlsx) sheets summarising tidally corrected transect time-series across 578 sites (e.g. nzd0058.xlsx); downstream analyses (e.g. change detection, mapping) can ingest these structured workbooks. Cell refs: notebook cells: not documented.