Abstract
Normalized Difference Vegetation Index (NDVI) is widely used as an efficient indicator of vegetation cover. Here we assess the possibility of using NDVI as an indicator of groundwater storage. We used groundwater level (GWL) obtained from in situ groundwater observation wells (n > 15,000) in India in 2005–2013. Good correlation (r > 0.6) is observed between NDVI and GWL in natural vegetation-covered areas, that is, forest lands, shrubs, and grasslands. We apply artificial neural network and support vector machine approaches to investigate the relationship between GWL and NDVI using both of the parameters as input. Artificial neural network- and support vector machine-simulated GWL matches very well with observed GWL, particularly in naturally vegetated areas. Thus, we interpret that NDVI may be used as a suitable indicator of groundwater storage conditions in certain areas where the water table is shallow and the vegetation is natural and where in situ groundwater observations are not available.
| Original language | English |
|---|---|
| Pages (from-to) | 8082-8092 |
| Number of pages | 11 |
| Journal | Geophysical Research Letters |
| Volume | 46 |
| Issue number | 14 |
| DOIs | |
| State | Published - 2019 |
| Externally published | Yes |
Funding
This manuscript uses freely available data of the Central Ground Water Board (CGWB), Government of India. We acknowledge CGWB, India, for providing water level data. We also acknowledge MODIS science team and TRMM satellite mission for providing NDVI and precipitation data, respectively. We acknowledge Dipankar Saha, CGWB, and AGI Project (IIT/SRIC/GG&CSE/AGI/2013-14/201) from Ministry of Human Resource Development (MHRD), Govt. of India.
| Funders | Funder number |
|---|---|
| Central Ground Water Board | |
| IIT/SRIC/GG | CSE/AGI/2013-14/201 |
| Ministry of Human Resource Development |
Keywords
- ANN
- Groundwater level prediction
- India
- NDVI
- SVM