# Vegetation Health Indicators of Groundwater Discharge: Integration of Sentinel-2 Remote Sensing and Meteorological Time Series in the Northern Apennines (Italy)

## TL;DR

This study shows that satellite data can detect areas where groundwater supports vegetation during droughts in mountainous regions.

## Contribution

A reproducible and cost-effective workflow using NDVI and SPEI to identify groundwater discharge zones in rugged landscapes.

## Key findings

- NDVI values in spring-centered areas are higher during droughts, indicating groundwater-supported vegetation resilience.
- Vegetation health anomalies can serve as reliable indicators of groundwater discharge likelihood in forested regions.
- The workflow is effective for hydrogeological reconnaissance in remote and mountainous environments.

## Abstract

What are the main findings?
The NDVI-derived vegetation anomalies indicate zones influenced by groundwater discharge within forested regions during droughts;Groundwater discharge areas enhance vegetation resilience under severe drought conditions in a Mediterranean mountain setting.

The NDVI-derived vegetation anomalies indicate zones influenced by groundwater discharge within forested regions during droughts;

Groundwater discharge areas enhance vegetation resilience under severe drought conditions in a Mediterranean mountain setting.

What are the implications of the main findings?
Remote sensing–based NDVI analysis can support reconnaissance and inventory of potential groundwater discharge zones in remote and rugged landscapes;Springs, including tapped springs used for human supply, represent key groundwater-dependent ecosystems in forested regions and require protection.

Remote sensing–based NDVI analysis can support reconnaissance and inventory of potential groundwater discharge zones in remote and rugged landscapes;

Springs, including tapped springs used for human supply, represent key groundwater-dependent ecosystems in forested regions and require protection.

This study evaluates the capability of multi-temporal vegetation indices derived from Sentinel-2 imagery to indicate groundwater discharge in a forested mountainous sector of the Northern Apennines (Italy). The NDVI was computed from Level-2A surface reflectance data (10 m resolution) and analyzed over five growing seasons (2017–2021), encompassing recurrent summer droughts. Aridity conditions were quantified using the Standardized Precipitation–Evapotranspiration Index (SPEI) derived from long-term meteorological records. The methodological framework integrates cloud-masked satellite observations, drought characterization, and spatial statistical comparison between known spring discharge zones and randomly distributed forested control points. NDVI values extracted within 100 m radius buffers, centered on spring outlets, were systematically compared with those from control areas located outside the shallow-water-table influence zone. During periods of negative SPEI (moderate-to-severe drought), spring-centered buffers consistently exhibited higher NDVI values than control sites, with the NDVI contrast increasing under severe arid conditions. This pattern indicates enhanced vegetation resilience supported by shallow groundwater availability. The results demonstrate that vegetation health anomalies, when constrained by homogeneous land cover and a consistent hydrogeological setting, can serve as indicators of the groundwater discharge likelihood. The proposed workflow provides a reproducible and cost-effective tool to support hydrogeological reconnaissance and spring inventorying in rugged mountainous environments where field-based surveys are logistically demanding.