How varunnaikr/satellite-soil-moisture-fusion-ml works
This is a small soil-moisture fusion script: it loads SMAP, Sentinel, and NDVI arrays, builds pixel-level features and a weighted target, trains and applies a gradient-boosting regressor, then visualizes and saves the fused array. The README provides no further capability details; the workflow is based on the single sampled source file.
Main entry points: fusion.py Fusion workflow · smap.npy SMAP input · sentinel.npy Sentinel input · ndvi.npy NDVI input
Components 10
Other components
- User
- Fusion workflow
fusion.py - SMAP input
smap.npy - Sentinel input
sentinel.npy - NDVI input
ndvi.npy - Features and target
fusion.py - Gradient boosting
- Fused array
fusion.py - Plots and comparisons
fusion.py - Fused output
fused.npy
Connections 10
- User → Fusion workflow: runs
- Fusion workflow → SMAP input: reads
- Fusion workflow → Sentinel input: reads
- Fusion workflow → NDVI input: reads
- Fusion workflow → Features and target: derives
- Features and target → Gradient boosting: trains
- Fusion workflow → Gradient boosting: predicts
- Gradient boosting → Fused array: returns predictions
- Fusion workflow → Plots and comparisons: plots
- Fusion workflow → Fused output: saves