Sentinel-1 and Sentinel-2 Data to Detect Irrigation Events: Riaza Irrigation District (Spain) Case Study

Abstract

Este estudio analiza el uso de mapas de humedad del suelo de alta resolución, obtenidos a partir de datos Sentinel-1 y Sentinel-2, para detectar ocurrencias de riego en tiempo y espacio en el distrito de riego de Riaza, España. Se emplea una base de datos detallada con información sobre uso del suelo, programación de riegos, consumos de agua, meteorología y límites parcelarios entre 2017 y 2021. Los resultados muestran que aproximadamente tres cuartas partes de los campos irrigados dentro de los tres días previos a la adquisición de Sentinel-1 pueden identificarse. La humedad del suelo permite detectar tempranamente los eventos de riego, incluso antes de que los índices de vegetación reflejen la presencia de cultivos, y posibilita precisar el momento del riego, algo que no es factible con dichos índices. Por ello, ambos enfoques son complementarios.

This paper investigates the use of high resolution (~100 m) surface soil moisture (SSM) maps to detect irrigation occurrences, in time and space. The SSM maps have been derived from time series of Copernicus Sentinel-1 (S-1) and Sentinel-2 (S-2) observations. The analysis focused on the Riaza irrigation district in the Castilla y León region (Spain), where detailed information on land use, irrigation scheduling, water withdrawal, meteorology and parcel borders is available from 2017 to 2021. The well-documented data basis has supported a solid characterization of the sources of uncertainties affecting the use of SSM to map and monitor irrigation events. The main factors affecting the irrigation detection are meteo-climatic condition, crop type, water supply and spatial and temporal resolution of Earth observation data. Results indicate that approximately three-quarters of the fields irrigated within three days of the S-1 acquisition can be detected. The specific contribution of SSM to irrigation monitoring consists of (i) an early detection, well before vegetation indexes can even detect the presence of a crop, and (ii) the identification of the irrigation event in time, which remains unfeasible for vegetation indexes. Therefore, SSM can integrate vegetation indexes to resolve the irrigation occurrences in time and space.