Data deposit and Images Article Title "Soil Management and Abandonment Detection in Mediterranean Olive Groves under Climate Stress: A Case Study from Central Spain"

This archive contains supplementary data and resources used in the analysis and modeling described in the manuscript. Below is a list of the files included, along with brief descriptions. 1. Raster Predictions of Soil Organic Carbon (SOC) and Brightness Index (BI) These GeoTIFF files represent Brightness Index (BI) and corresponding predicted SOC values over time, derived using remote sensing. **SOC Prediction:** - 2015prediction.tif - 2018prediction.tif - 2021prediction.tif - 2024prediction.tif **Brightness Index (BI):** - BI2015decimal.tif - BI2018decimal.tif - BI2021decimal.tif - BI2024decimal.tif Format: GeoTIFF (.tif) Resolution: 30 m Each `.tif` is accompanied by a `.tif.aux.xml` metadata file automatically generated by GIS software. 2. Field and Pixel Data - **Alcocer six points.xlsx** Field SOC measurements and management metadata from 6 ground-truth sampling points in Alcocer, with corresponding remote sensing indices. - **Bergonza data.xlsx** Field SOC measurements and management metadata from 60 ground-truth sampling points in Bergonza, with corresponding remote sensing indices. - **Pixel center 30 data.xlsx** Pixel-level SOC prediction dataset (30 m resolution), derived using Equation (3) in the manuscript. 3. Object Detection Dataset - **Abandonment-Spain-8** *(folder)* This folder contains the dataset used for training and validation of olive grove abandonment detection models (YOLO). Includes: - Dataset split (train / test / validation) - Corresponding label files Refer to the main manuscript for detailed methodology, model architecture, and interpretation guidelines. Corresponding author: Giovanni Marchese