Orbit to Ground · NASA International Space Apps Challenge 2026 · Be An Earth System Trend Detective! · Team GeoAI Toronto (Parsa Riahi). Source code: github.com/Priahi/nasa-spaceapps-2026 (Apache-2.0).
Data: NASA FIRMS, NASA-ISRO NISAR (via NASA CMR and ASF DAAC), NASA POWER, NASADEM, MODIS, plus ESA Copernicus Sentinel-1, Sen1Floods11, USDA NASS CDL and ISRIC SoilGrids. Province outlines: Natural Earth (public domain). Each project page lists its sources and licences. This site is not affiliated with or endorsed by NASA.
AI disclosure: AI coding assistants helped write and refactor parts of the code, site and text. All models, metrics and figures come from our own runs on the public data listed. Model metrics are one seed each, not a published benchmark. Built on open-source libraries (pandas, NumPy, LightGBM, PyTorch, ONNX Runtime, pymannkendall, Leaflet, GSAP, Lenis) and our own earlier code: the three tools started before the event, and the trend layer was computed during it (14 Nov 2026).