@article{fdi:010085974,
  title = {{E}valuating downscaling methods of {GRACE} ({G}ravity {R}ecovery and {C}limate {E}xperiment) data : a case study over a fractured crystalline aquifer in southern {I}ndia},
  author = {{P}ascal, {C}. and {F}errant, {S}ylvain and {S}elles, {A}. and {M}arechal, {J}. {C}. and {P}aswan, {A}. and {M}erlin, {O}.},
  editor = {},
  language = {{ENG}},
  abstract = {{GRACE} ({G}ravity {R}ecovery and {C}limate {E}xperiment) and its follow-on mission have provided since 2002 monthly anomalies of total water storage ({TWS}), which are very relevant to assess the evolution of groundwater storage ({GWS}) at global and regional scales. {H}owever, the use of {GRACE} data for groundwater irrigation management is limited by their coarse (similar or equal to 300 km) resolution. {T}he last decade has thus seen numerous attempts to downscale {GRACE} data at higher - typically several tens of kilometres - resolution and to compare the downscaled {GWS} data with in situ measurements. {S}uch comparison has been classically made in time, offering an estimate of the static performance of downscaling (classic validation). {T}he point is that the performance of {GWS} downscaling methods may vary in time due to changes in the dominant hydrological processes through the seasons. {T}o fill the gap, this study investigates the dynamic performance of {GWS} downscaling by developing a new metric for estimating the downscaling gain (new validation) against non-downscaled {GWS}. {T}he new validation approach is tested over a 113 000 km(2) fractured granitic aquifer in southern {I}ndia. {GRACE} {TWS} data are downscaled at 0.5 degrees (similar or equal to 50 km) resolution with a data-driven method based on random forest. {T}he downscaling performance is evaluated by comparing the downscaled versus in situ {GWS} data over a total of 38 pixels at 0.5 degrees resolution. {T}he spatial mean of the temporal {P}earson correlation coefficient ({R}) and the root mean square error ({RMSE}) are 0.79 and 7.9 cm respectively (classic validation). {C}onfronting the downscaled results with the non-downscaling case indicates that the downscaling method allows a general improvement in terms of temporal agreement with in situ measurements ({R}=0.76 and {RMSE} = 8.2 cm for the non-downscaling case). {H}owever, the downscaling gain (new validation) is not static. {T}he mean downscaling gain in {R} is about +30 % or larger from {A}ugust to {M}arch, including both the wet and dry (irrigated) agricultural seasons, and falls to about +10 % from {A}pril to {J}uly during a transition period including the driest months ({A}pril-{M}ay) and the beginning of monsoon ({J}une-{J}uly). {T}he new validation approach hence offers for the first time a standardized and comprehensive framework to interpret spatially and temporally the quality and uncertainty of the downscaled {GRACE}-derived {GWS} products, supporting future efforts in {GRACE} downscaling methods in various hydrological contexts.},
  keywords = {{INDE} ; {TELANGANA}},
  booktitle = {},
  journal = {{H}ydrology and {E}arth {S}ystem {S}ciences},
  volume = {26},
  numero = {15},
  pages = {4169--4186},
  ISSN = {1027-5606},
  year = {2022},
  DOI = {10.5194/hess-26-4169-2022},
  URL = {https://www.documentation.ird.fr/hor/fdi:010085974},
}