Water bodies recognition for depression-focused recharge. Tambov region, Russia

Published: 31 August 2021| Version 1 | DOI: 10.17632/pn3gdzhdy4.1


In the semiarid ecoregion of temperate zone focused snowmelt water infiltration in the topographic depressions is a key groundwater recharge mechanism. In this paper, remote sensing was used to construct a mass balance and estimate the volume of the ephemeral ponds from the time series of high-resolution Planet Labs images taken in April to May 2021 in Tambov area, Russia. First small water bodies were automatically recognized on each image separately using object-based super-vised classification. With the use of defined water pixels obtained from a small unmanned aerial vehicle on one of the latest images as a training set, it was possible to get a prediction for the earlier image with =0.99. The DEM was used to estimate the water volume that decreased according to the negative exponential equation with time and the power of this exponent did not systematically depend on the pond size. Function-based interpolation of the water bodies area and volume and daily estimation of Penman evaporation allowed to calculate daily infiltration into the depression beds as a reminder. The infiltration was maximum at the spring onset (5-40 mm/day) and decreased with time. Obtained spatially variable infiltration rate has improved the steady-state shallow groundwater simulation.


Steps to reproduce

1. Download high-resolution chronological orthophoto maps in the visible range (3 meters) from the sensors of the RapidEye and SkySat minisatellites of the Planet Labs system 2. Use he Interactive Supervised Classification tool in the ArcGIS Pro desktop application 3. Conbine with available DEM to estimate the volume of ponds * 4. Use the python code to esmimate infiltration from the pond volume and area *DEM is available on request due to country restriction


Pochvenniy Institut imeni V V Dokuchaeva


Hydrology, Remote Sensing