Saved in:
Bibliographic Details
Main Author: Verma, Divyam
Format: Recurso digital
Language:English
Published: Zenodo 2026
Subjects:
Online Access:https://doi.org/10.5281/zenodo.20259081
Tags: Add Tag
No Tags, Be the first to tag this record!
Table of Contents:
  • <div> <div>This work presents an end-to-end pipeline for spatio-temporal</div> <div>  crop health assessment in the Jammu region of India. Multi-</div> <div>  source Earth observation data Sentinel-2 surface reflectance,</div> <div>  MODIS land-surface temperature, and CHIRPS precipitation are</div> <div>  harmonised over 150 fixed sampling points to derive seven</div> <div>  vegetation and hydro-climatic features. A composite Crop Health</div> <div>  Score is modelled with classical and deep-learning regressors</div> <div>  under a strict temporal hold-out, interpreted via SHAP, and</div> <div>  extended to recursive multi-step forecasting. A Streamlit</div> <div>  application exposes the trained model for on-demand,</div> <div>  coordinate-level prediction.</div> </div>