Evaluation of GRACE derived groundwater storage changes in different agro-ecological zones of the Indus Basin

Highlights

• The GRACE’s performance was evaluated in two different agro-ecological zones.

• The trend of groundwater abstraction was analyzed from 2002 to 2017.

• The correlation of GRACE with in-situ measurements was better at plain regions.

Abstract

The Gravity Recovery and Climate Experiment (GRACE) has recently been identified as a useful tool for monitoring changes in groundwater storage (GWS), especially in areas with sparse groundwater monitoring networks. However, GRACE's performance has not been evaluated in the highly heterogeneous Indus Basin (IB) to date. The objective of this study was thus (i) to evaluate GRACE's performance in two distinctively different agro-ecological zones of the IB, and (ii) to quantify the trend of groundwater abstraction over 15 years (i.e., from 2002 to 2017). To capture this heterogeneity at the IB, the two different agro-ecological zones were selected: i) the Kabul River Basin (KRB), Afghanistan, and ii) the Lower Bari Doab Canal (LBDC) command area in Pakistan. The groundwater storage anomalies (GWSA) for both regions were extracted from random pixels. The results show a correlation (R²) of 0.46 for LBDC and 0.32 for the KRB, between the GWSA and in-situ measurements. The results further reveal a mean annual depletion in GWSA of −304.2 ± 749 and −301 ± 527 mm at the LBDC and the KRB, respectively. Overall, a net GWS depletion during 2002–2017 at the LBDC and KRB was 4.87 and 4.82 m, respectively. The GWSA’s response to precipitation analyzed through cross-correlation shows a lag of 4 and 3 months at the KRB and the LBDC, respectively. The GWSA's poor correlation with the in-situ measurements particularly in the mountainous region of the KRB is driven by the 4 months lag time unlike in the LBDC (i.e. 3 months); besides, the observations wells are sparse and limited. The complex geomorphology and slope of the landscape also cause discrepancies in the correlation of the in-situ measurements and the GRACE-derived changes in GWS at the two different agroecological zones of the IB. The spatially averaged GWSA in monthly time steps is another reason for the lower correlation between GRACE-based GWSA estimates and point-based in-situ measurements. Therefore, care must be taken while using GRACE's output in regions with heterogeneous geomorphologic features.

Introduction

Groundwater use provides distinct benefits and chances for human development (Velis et al. 2017). It is approachable to a wide number of consumers (Sakthivadivel 2007) and can be utilized based on its depth, the reason for which it is extracted, the relationship between recharge rate against withdrawals, and the distance between wells. When not tainted by human interference, the main advantage of groundwater is its high microbiological purity, which stems from its underground location and the natural protection by the soil and geological layers and filtration during flow (Calow et al. 1997). Globally, the rate of groundwater depletion rose from 126 ± 32 km³ year⁻¹ to 283 ± 40 km³ year⁻¹ between 1960 and 2000, respectively (Wada et al. 2010). According to Zektser and Everett (2004), 20% of the global irrigation withdrawals, around 40% of total commercial activities, and 50% of total urban activities are supported with withdrawals from groundwater. Thus, depletion in groundwater resources will cause a significant interruption of their societal and ecological functions (Danielopol et al., 2003). In both Afghanistan and Pakistan, the population’s reliance on groundwater abstraction for municipal, agricultural, and industrial purposes is growing day by day. As a result, groundwater levels are falling, prompting the consumers in both countries to respond sustainably and adaptively. There is a massive withdrawal of groundwater in the region of the Indus Basin (IB); according to Uhl (2006), the Kabul River Basin (KRB) experiences a withdrawal of 450 million m³ against recharge of 380 million m³. Similarly, Pakistan is experiencing a mean annual recharge of 83.88 km³ against a groundwater withdrawal of 55 km³ (Bhutta 2005)_.

More comprehensive groundwater governance is required to ensure the protection of groundwater resources. This should include spatial resolution mapping of groundwater aimed at detecting changes in water level and volume, which has largely been overlooked thus far (Megdal et al. 2015). To effectively estimate the groundwater resources and associated fluctuations, it is necessary to establish an effective groundwater-monitoring network (Wu 2004). There are various tools and methodologies that could be used at different spatial and temporal scales to assess the changes in groundwater. Usually, the groundwater observation wells are mostly used to indirectly extract the groundwater use information. It gives point-based measurements of water levels to higher accuracy, and so far the observation wells, both in Afghanistan and Pakistan, are sparsely installed and their spatial and temporal maintenance and monitoring have not been systematically managed either. In Pakistan, the groundwater observations are taken twice a year, e.g. pre-and post-monsoon season (Shahzad et al. 2020). The systematic errors may be avoided but measurements must be performed exclusively by personnel who have received adequate training (Rau et al. 2019). In addition to systematic errors in the measurements, the representativeness of the site selection for the monitoring points is a very important issue to minimize systematic errors. For example, a location close to main irrigation canals or drainage collectors might systematically influence or disturb the representativeness of measurements. Yet, this approach is practically not cost-effective when dealing with large river basins, countries, or continental geographies. Instead, there is a need to use a system/tool/method which is less expensive, yet accurate and capable to cover larger spatial extents with appropriate time-steps. Therefore, accurate and periodic long-term measurements of aquifer levels are important to track the storage of groundwater and its changes over space and time (Taylor and Alley 2001).

Recently, Gravity Recovery and Climate Experiment (GRACE) based groundwater storage (GWS) variation, both in space and time and in complex aquifer systems in large heterogeneous river basins has been very instructive (Singh and Saravanan, 2020, Iqbal et al., 2016). The GRACE satellite project provides new insights into mass redistribution within the Earth structure and establishes new hydrological viewpoints (Tapley et al., 2004a, Tapley et al., 2004b). The GRACE satellite tracks Spatio-temporal changes with an unparalleled resolution in the Earth’s gravity field and facilitates the analysis of mass changes within hydro-systems. The detailed estimation of the GWS changes derived from the GRACE satellite could be used for a decision support system in maintaining the recharge and discharge strategies within a catchment or basin. Since GRACE measures terrestrial water storage (TWS), deriving GWS change out of the TWS column requires to be validated with in-situ measurements of groundwater levels before being used for water resources management initiatives under local circumstances (Neves et al., 2020, Zhong et al., 2018, Yin et al., 2018).

The objective of this study was thus (i) to evaluate the GRACE’s performance in two different agro-ecological zones of the IB, as well as (ii) to quantify the trend of groundwater abstraction over 15 years (i.e., 2002–2017). The KRB (in Afghanistan) and the Lower Bari Doab Canal (LBDC) command area (in Pakistan) were the two distinct agro-ecological zones that were rigorously examined under the auspices of this study. The KRB and LBDC distinctly differ from each other not only in terms of geomorphologic features but also in terms of having large changes in the irrigation infrastructure, elevation, slope, climatic conditions as well as management hierarchies that undergo transboundary dissociated up-and-downstream interactions too. The addition of such an agro-ecological heterogeneity component to the evaluation of GRACE’s performance is an innovative part of this study concerning previous research in the two regions. The detailed Spatio-temporal estimates of GRACE-based GWS were also validated against in-situ groundwater levels; Furthermore, groundwater storage anomalies (GWSA) were cross-correlated with precipitation data to determine how variations in GWS respond to precipitation.