Estimation of the water table based on NDVI and NDWI Landsat, in the South Citrus Aquifer: a case study

Abstract

Objective: to assess the spatiotemporal changes of the vegetation of the Southern Citrus Aquifer and their possible association with the depth of the water table, through the analysis of the vegetation cover and the piezometric level with geomatics on Landsat images and the NDVI and NDWI spectral indices.

Design/ Methodology/ Approach: groundwater depth level data were collected from 63 monitoring wells distributed in the Southern Citrus Aquifer (DR-1914). And representative Landsat images corresponding to the years 2003, 2013 and 2022 (Landsat TM for 2003 and 2013, and Landsat OLI for 2022) were processed and analyzed with geomatics. The spatial distribution of the water table for each year was estimated by means of geostatistical interpolation techniques and the information was processed with QGIS^© 3.28 and Google Earth Engine^©.

Results: our results showed a general trend of increase in NDVI values between 2003 and 2013; which changed towards signs of water stress in 2022. This means that there are changes in vegetation cover and vigor in relation to water availability and local anthropogenic pressure. The depth of the water table in 2022 differs significantly (p ≤ 0.001) from the same value in 2003. This supports the existence of a progressive aquifer depletion process that spectral indices can detect, to assess the underground dynamics of the plant systems associated with the aquifer.

Limitations/ Implications of the study: discontinuity was identified in the piezometric records, due to the absence of measurements in some years. This limitation was mitigated by interpolation estimation of average values from earlier and later records.

Findings/ Conclusions: NDVI and NDWI indices proved to be effective tools for analyzing the relationship between vegetation dynamics, soil moisture availability, and water table depth. Remote sensing offers a methodological alternative for monitoring the hydrological and environmental dynamics of the aquifer. This methodology is efficient, spatially continuous, and temporally consistent.