Saswata Nandi

Search

Saswata Nandi

Postdoctoral Researcher · HydrologyUniversity of California, MercedCalifornia, USA

I study how a warming climate, wildfire and changing forests alter water availability and the risk of floods and droughts. My approach pairs land-surface modelling with satellite observations and climate records. I completed my PhD at IIT Bombay, where I developed IMDLIB, an open-source Python library that is now a widely used tool for working with India’s long-term gridded rainfall and temperature records.

  • CV
  • Scholar
  • ORCID
  • GitHub
  • Email
Photo of Saswata Nandi

Research

All research

News

  1. Released IMDLIB 0.3, with one function to load IMD gridded data and named regions of India (states, districts, basins and cities). Read

  2. New paper in Frontiers in Water on crop water use in the San Joaquin Valley, and how much water other land uses could save. Read

  3. At AGU Fall Meeting 2025 with work on how drought affects ecosystem water-use efficiency across the US (with A. Chakraborty, J. Abatzoglou and M. Safeeq).

  4. New paper in Journal of Water and Climate Change on how climate change may affect runoff and drought risk in the Upper Bhima basin. Read

  5. New paper in Ecohydrology on how forest management changes the water balance of the upper Kings River Basin. Read

Selected papers

All publications

Comparing crop evapotranspiration estimation methods to evaluate water-saving potential of alternative land uses in the San Joaquin Valley, California

Kandhway, A., Nandi, S., Conklin, M. H. & Safeeq, M.

Frontiers in Water, 8, 1879797

DOI

Software

All software
  • Python
  • MIT license
  • v0.3.1

IMDLIB

Open-source Python package for retrieval, processing, and spatiotemporal analysis of gridded India Meteorological Department (IMD) observational datasets.

284,147 PyPI downloadsas of 28 Sept 2026

Repository Documentation PyPI Paper

More about IMDLIB 

$ pip install imdlib

# daily rainfall over India, 2020 to 2023
>>> import imdlib as imd
>>> data = imd.load('rain', 2020, 2023)
>>> ds = data.get_xarray()
>>> ds['rain'].mean('time').plot()

Blog and reading

Latest postWhat's new in IMDLIB 0.3The new features of IMDLIB 0.3, simpler loading and built-in regions of India, shown with IMD data. Recently readA manifesto for the equifinality thesisWhy a good fit to streamflow does not mean we have found the right model.Beven, K. · Journal of Hydrology · 2006 · read Oct 2026