Análisis de la dinámica espacio-temporal de las lagunas del Parque Nacional Cajas usando datos radar Sentinel-1

Abstract

This study analyzed the spatiotemporal dynamics of the lagoons in Cajas National Park, southern Ecuador, using Sentinel-1 radar imagery, which provides information independently of the cloud cover that characterizes the area. The research was conducted using a multitemporal series of 20 images corresponding to the 2017–2026 period, considering both wet and dry seasons. The images underwent a preprocessing workflow that included study area clipping, precise orbit application, thermal noise removal, radiometric calibration, speckle noise filtering, and terrain correction to ensure data quality. Subsequently, a segmentation technique based on the manual thresholding of backscatter values, supported by their statistical analysis, was applied to delineate water bodies and generate a spatiotemporal database. The results revealed variations in the total cumulative surface area of the 100 lagoons studied, with values ranging from 915 ha to 930 ha during the wet season, while in the dry season they ranged from 683 ha to 747 ha, showing interannual fluctuations characterized by alternating periods of increase, decrease, and relative stability in both seasons. Furthermore, it was observed that, following certain periods of reduction in water surface area, the lagoons did not always fully recover to previously recorded levels, reflecting variable dynamics throughout the study period.