Estimation of Groundwater Storage Variations in Indus River Basin Using Grace Data

Yahya Sattar, Zubair Khalid

TL;DR

Researchers have developed a more accurate way to track groundwater depletion in the Indus River Basin (IRB) by combining NASA’s GRACE satellite gravity data with GLDAS hydrological models. The core innovation is a specialized "optimal window" mathematical function that precisely isolates the IRB region, minimizing the data "leakage" errors that typically plague large-scale satellite water measurements.

What Problem Does This Paper Address?

Monitoring groundwater is a massive logistical challenge. Traditional methods rely on physical wells, which are expensive, slow to report, and provide "point" data that doesn't always represent a whole region. This is especially critical in the Indus River Basin, where over 60% of water is pumped from the ground to support millions of people, often with little oversight.

While satellites like GRACE (Gravity Recovery and Climate Experiment) can detect changes in water mass from space, the data is delivered as "spherical harmonics"—a complex mathematical format. When researchers try to "zoom in" on a specific area like the IRB using standard methods (like a simple boxcar/rectangular filter), they encounter significant errors because the signal "leaks" in from surrounding regions. This paper aims to fix that "leakage" to provide a more reliable estimate of groundwater storage (GWS) variations.

Methods

The authors propose a data processing framework that separates Groundwater Storage (GWS) from Total Water Storage (TWS).

  1. The TWS Calculation: They use GRACE gravity coefficients to calculate the total change in mass over the Earth's surface. This includes everything: snow, soil moisture, surface water, and groundwater.
  2. The "Optimal Window" (The Secret Sauce): Instead of just cutting out the shape of the Indus River Basin (which creates mathematical artifacts), they design a band-limited optimally concentrated window function. Using Slepian spatial-spectral concentration theory, they find a mathematical "window" that is perfectly tuned to the IRB's shape. To make this computationally feasible, they approximated the IRB's irregular shape using 44 smaller sub-regions.
  3. The Subtraction Method: To isolate groundwater, they use the Global Land Data Assimilation System (GLDAS) to estimate and subtract the "noise"—specifically Soil Moisture Storage (SMS) and Snow Water Equivalent (SWE)—from the total water signal.
  4. Smoothing: They apply a von Mises function (a spherical version of Gaussian smoothing) to reduce high-frequency noise in the satellite data.

Datasets

  • GRACE (Release 03): Monthly gravity solutions providing Stokes coefficients used to calculate mass changes.
  • GLDAS: A high-resolution ($1/4^\circ$) land surface model used to estimate soil moisture and snow levels.

Key Insights and Ideas

  • The Boxcar Pitfall: The paper highlights that standard spatial averaging is equivalent to a "boxcar window." In the world of spherical harmonics, a boxcar has an infinite band-limit, meaning it inevitably introduces inaccuracies when trying to localize data.
  • Slepian Concentration: By treating the localization as an eigenvalue problem, the researchers can find the "eigenfunction" that puts the maximum amount of signal energy exactly where the IRB is, significantly improving the signal-to-noise ratio for that specific basin.
  • Analytic Efficiency: By breaking the IRB into 44 "limited colatitude-longitude" regions, the authors were able to use exact analytical expressions to solve the complex integrals required for the window design, rather than relying on slower numerical approximations.

Key Results

  • Seasonal Trends: The framework successfully captured the expected seasonal variations in the Indus River Basin (recharging during monsoons/snowmelt and depleting during dry seasons).
  • Long-term Depletion: The analysis of data from 2005 to 2015 revealed a clear downward trend in groundwater levels. The linear fitting of the processed data confirms that the IRB is experiencing steady groundwater depletion over time.
  • Methodological Validation: The "optimal window" approach proved effective at localizing the signal within the complex geometry of the IRB, providing a blueprint for monitoring other critical basins globally.

Limitations and Future Work

  • Dynamic Oblateness: The authors note that they did not account for significant variations in the Earth’s "dynamic oblateness" (the flattening at the poles), which could affect gravity readings.
  • Data Sources: The study suggests that incorporating data from ESA’s GOCE (Gravity field and steady-state Ocean Circulation Explorer) mission could further refine the results.
  • Window Refinement: Future research may explore using a "weighted sum" of multiple eigenfunctions for the window design, rather than just the single most concentrated one, to potentially capture even more spatial detail.

Summary updated

Abstract

The depletion and variations of groundwater storage (GWS) are of critical importance for sustainable groundwater management. In this work, we use Gravity Recovery and Climate Experiment (GRACE) to estimate variations in the terrestrial water storage (TWS) and use it in conjunction with the Global Land Data Assimilation System (GLDAS) data to extract GWS variations over time for Indus river basin (IRB). We present a data processing framework that processes and combines these data-sets to provide an estimate of GWS changes. We also present the design of a band-limited optimally concentrated window function for spatial localization of the data in the region of interest. We construct the so-called optimal window for the IRB region and use it in our processing framework to analyze the GWS variations from 2005 to 2015. Our analysis reveals the expected seasonal variations in GWS and signifies groundwater depletion on average over the time period. Our proposed processing framework can be used to analyze spatio-temporal variations in TWS and GWS for any region of interest.
BibTeX
@inproceedings{sattar2021estimation,
  author = {Yahya Sattar and Zubair Khalid},
  title = {{Estimation of Groundwater Storage Variations in Indus River Basin Using Grace Data}},
  booktitle = {IEEE International Conference on Acoustics, Speech and Signal Processing},
  year = {2021},
  url = {https://openreview.net/forum?id=vLFEn0WBmf},
}

Loading related content...

Figure