Thu Oct 08, F. Flechtner
• Sorted by Date • Sorted by Last Name of First Author •
Bhunia, Partha Sarathi and Abhishek, 2027. On the improved reconstruction of GRACE–TWSA for assessing droughts and their governing factors in a dam–regulated region. Physics and Chemistry of the Earth, 145:104765, doi:10.1016/j.pce.2026.104765.
• from the NASA Astrophysics Data System • by the DOI System •
@ARTICLE{2027PCE...14504765B,
author = {{Bhunia}, Partha Sarathi and {Abhishek}},
title = "{On the improved reconstruction of GRACE-TWSA for assessing droughts and their governing factors in a dam-regulated region}",
journal = {Physics and Chemistry of the Earth},
keywords = {GRACE, Drought, Deccan plateau, Terrestrial water storage, Anthropogenic activities},
year = 2027,
month = jan,
volume = {145},
eid = {104765},
pages = {104765},
abstract = "{Accurate drought assessment in heavily managed basins requires
continuous terrestrial water storage anomaly (TWSA) records that
capture both climatic and anthropogenic signals. This study asks
whether data-driven reconstructions of GRACE/GRACE-FO TWSA
require complex models with large predictor sets, and how
different approaches perform in regions with intensive human
activity. We use a classical artificial neural network (ANN),
forced with a targeted set of six predictors, including two
storage components that are strongly influenced by anthropogenic
interventions over the dam-regulated Deccan Plateau, India. ANN-
based reconstructed TWSA yields high skill across 11
meteorological subdivisions (correlation = 0.96--0.99, NSE =
0.80--0.99) and performs comparably to, or better than, existing
global reconstruction products, particularly within heavily
managed regions. Experiments excluding SWS or GWS increase RMSE
by up to two-fold, demonstrating that explicitly representing
these anthropogenically sensitive storage components is critical
for improving TWSA reconstruction in such managed basins.
Drought analysis using GRACE-Drought Severity Index (DSI) and
Combined Climatologic Deviation Index (CCDI) reveals hotspots
and extremes: the maximum number of events (seven) occurs in
Marathwada and South Interior Karnataka, while the 2015--2019
event in North Interior Karnataka records the largest duration
(49 months) and severity (CCDI = -52.99; GRACE-DSI = -56.69).
When assessing drought-induced water losses for 2003--2016, our
product shows the closest agreement with GRACE, outperforming
alternative datasets. Among the four large-scale climate
oscillations examined, ENSO is the dominant mode associated with
drought variability. Overall, the findings indicate that
physically informed predictor selection is more important than
increasing model complexity or predictor quantity for improving
regional TWSA reconstruction in heavily managed basins,
highlighting the value of incorporating anthropogenically
sensitive storage components for drought assessment and water-
resources management.}",
doi = {10.1016/j.pce.2026.104765},
adsurl = {https://ui.adsabs.harvard.edu/abs/2027PCE...14504765B},
adsnote = {Provided by the SAO/NASA Astrophysics Data System}
}
Generated by
bib2html_grace.pl
(written by Patrick Riley
modified for this page by Volker Klemann) on
Thu Oct 08, 2026 16:44:00
GRACE-FO
Thu Oct 08, F. Flechtner