garrettgcraig/sparcle_ssa_agfo_annual_water_yield

Geospatial data pipeline for InVEST Annual Water Yield modeling across sub-Saharan Africa—preparing present-day and SSP3-70 climate projection inputs under the SPARCLE framework. # SPARCLE Ecosystem Services - Annual Water Yield Geospatial data processing pipeline for modeling the impact of agroforestry on annual water yield across sub-Saharan Africa using InVEST (Integrated Valuation of Ecosystem Services and Tradeoffs). The pipeline prepares present-day and future (SSP3-7.0) climate, land-cover, and soil input layers for the InVEST Annual Water Yield model, runs the model across every watershed × climate × agroforestry-scenario combination, and quantifies the resulting change in water yield. ## Study Area The **Afrotropic realm** of sub-Saharan Africa. The study area is the dissolved union of all HydroBASINS level-06 watersheds that intersect any Afrotropic ecoregion (RESOLVE Ecoregions 2017); whole watershed polygons are retained even where they only partially overlap the realm, because InVEST operates on complete watersheds. The boundary is built reproducibly in `scripts/_study_area.R` (`load_study_area()`). All spatial layers are projected to **ESRI:102022** (Africa Albers Equal Area Conic). ## Workflow All numbered scripts live in `scripts/`. ### 1. Land Use / Land Cover (`01_SPARCLE_LULC.ipynb`) Downloads ESA WorldCover 2020 (10 m) land cover data for Western Africa via Google Earth Engine, aggregated to 100 m and exported as GeoTIFF tiles. > **Urban (GAO) input.** The mid-century urban-expansion layer is **not** generated by a GEE notebook. It comes from the Gao & Pesaresi (2021) SEDAC product and is consumed directly by step 8a — see step 8a and the Obtaining the input data table. (A retired GEE notebook, `scripts/archive/01b_GAO_urban_LULC_ssp370.ipynb`, fetched a different Gao & O'Neill product and is no longer part of the pipeline.) ### 2. MODIS LAI Climatology (`02_SPARCLE_LAI.ipynb`) Downloads MODIS MOD15A2H v6.1 LAI (500 m) via Google Earth Engine and computes a monthly climatology (Jan--Dec) averaged across 2002--2025. Exports 12 GeoTIFFs to Google Drive for use in the Kc workflow. ### 3. Input Layer Assembly (`03_ …