---

name: timelapse-and-apps description: "Create geemap timelapse animations, local GIF/MP4 post-processing, and app export/deployment workflows." disable-model-invocation: true metadata: disco-role: operating license: MIT

geemap Timelapse and Apps

Use this sub-skill when a task asks for geemap satellite timelapse generation, animated GIF or MP4 post-processing, map-to-app export, or lightweight publication/deployment patterns.

Start here

1. Confirm geemap, Earth Engine, and optional app dependencies with ../../references/installation-and-auth.md.
2. Decide whether the task is local-only or remote Earth Engine:
   • Local-only GIF/MP4 annotation/conversion: use scripts/gif_tool.py and references/workflows.md.
   • Remote satellite timelapse generation: use references/workflows.md and the function notes in references/api-reference.md.
   • Web/app output: use references/workflows.md.
3. If anything fails, check references/troubleshooting.md before changing the workflow.

What this sub-skill owns

• Earth Engine timelapse functions from geemap.timelapse: Landsat, Sentinel-2, Sentinel-1, GOES, NAIP, MODIS NDVI/ocean-color, generic ImageCollection timelapse, Dynamic World timeseries, text/progress overlays, fading, and GIF-to-MP4 conversion.
• geemap.Map.add_landsat_ts_gif when the user wants to draw/select an ROI on an interactive map and overlay the generated Landsat GIF back onto the map.
• Deterministic local post-processing of an existing GIF: inspect frames, create tiny fixture GIFs, add frame labels/progress bars through geemap, and convert to MP4 when ffmpeg is available.
• Map output and app routes: Map.to_html, Map.to_streamlit, folium to_gradio, folium Streamlit bidirectional mode, Voila/ngrok-style notebook serving, Solara/Streamlit/Gradio optional extras, and folium Datapane publishing when the required package and token are present.

Route elsewhere

• General Earth Engine layer creation, basemaps, widgets, draw controls, inspectors, or map backend selection: ../interactive-earth-engine-maps/SKILL.md.
• Static maps, charts, colorbars/legends outside an animation, cartoee outputs, Plotly/pydeck/kepler/maplibre visualization decisions: ../visualization-and-charts/SKILL.md.
• Generic Earth Engine exports, shapefile/GeoJSON/CSV/raster conversion, COG/STAC/OSM data movement, or export task monitoring not tied to a timelapse output: ../conversion-and-io/SKILL.md.
• ML classifiers, Dynamic World classification analysis beyond animation/time-series display, Gemini/AI dataset discovery, or classifier conversion: ../machine-learning-and-ai/SKILL.md.

Operating constraints

• Keep local GIF/MP4 processing separate from remote Earth Engine generation. Local processing should not initialize Earth Engine or require credentials.
• Remote timelapse generation requires an initialized Earth Engine session, network access, a valid ROI, sensible date ranges, and output dimensions small enough for Earth Engine video/thumbnail services.
• Most timelapse functions write a GIF path and may also write an MP4 when mp4=True; MP4 conversion requires a system ffmpeg binary.
• Large exports are more fragile. Start with a small ROI, dimensions<=768, annual or monthly frequency, and a short date span before increasing size or temporal density.
• App routes require optional extras and service tokens depending on target. Treat Streamlit, Gradio, Voila, Solara, Datapane, ngrok, and browser ports as deployment prerequisites, not core geemap computation.
• Do not rely on source notebooks or repository checkout files at runtime; the self-contained references and bundled script in this sub-skill are the supported operating material.