Key natural influences on groundwater storage changes in central and southern Arizona

Key natural influences on groundwater storage changes in central and southern Arizona | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Key natural influences on groundwater storage changes in central and southern Arizona Behshad Mohajer, James S. Famiglietti, Hrishikesh A. Chandanpurkar, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7698407/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 24 Mar, 2026 Read the published version in Scientific Reports → Version 1 posted 10 You are reading this latest preprint version Abstract Groundwater resources in central and southern Arizona are increasingly threatened by prolonged drought and rising temperature, curtailed surface water supply, and pumping overdraft. Recent assessments indicate that groundwater storage in the region is rapidly declining. While human influences are well-understood, natural hydroclimatic controls on groundwater storage changes are examined in less detail. Using NASA's Gravity Recovery and Climate Experiment (GRACE) and its successor GRACE-Follow On (GRACE-FO) Groundwater Storage Anomalies (GWSA) and land-surface variables from Western Land Data Assimilation System (WLDAS), we quantified standardized trends, subbasin-scale correlations, and dominant modes via Principal Component Analysis (PCA) from 2004 to 2021 in central and southern Arizona. Using key components, we grouped groundwater subbasins using k-means clustering to simplify assessments. Results show substantial spatial heterogeneity arising from recharge-responsive northern and central subbasins linked to precipitation and subsurface runoff, and southern loss-dominated subbasins linked to weaker recharge coupling and more substantial influence from atmospheric demand. PCA analysis shows that natural modes variability explains about 16% of the spatial variance displayed in GRACE/FO GWSA in the study area, with approximately 72% of that variance explained by total evapotranspiration (≈ 29%), precipitation (≈ 23%), and subsurface runoff (≈ 20%). The remaining variance reflects a combination of anthropogenic influences (pumping, MAR), geologic heterogeneity, and residual model/measurement errors. Our diagnostic framework presents groundwater subbasin scale clusters driven by shared hydroclimatic modes of variability. It has potential as a transferable tool for focused recharge, groundwater sustainability plans, and post-2026 groundwater planning in the Lower Colorado River Basin. Earth and environmental sciences/Environmental sciences Earth and environmental sciences/Hydrology Earth and environmental sciences/Solid earth sciences GRACE satellites WLDAS Principal Component Analysis Active Management Areas Lower Colorado River Basin Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 1. Introduction The Colorado River Basin (CRB) encompasses nearly 250,000 square miles of the arid and semi-arid American Southwest, draining portions of seven western US states and northern Mexico [ 1 ]. The CRB is hydrologically divided by the Lees Ferry, Arizona. The Upper Colorado River Basin (UCRB) delivers approximately 90% of the total flow into the CRB. At the same time, the Lower Colorado River Basin (LCRB) of Arizona, California, and Nevada, which is home to most of the urban and agricultural areas, contributes over 55% of the consumptive use [ 2 ]. This major river system supports over 40 million people, irrigates over five million acres of farmland, and supports a diverse set of ecosystems and economies in the region [ 3 ]. The UCRB encompasses a unique geological setting of fractured sedimentary formations with confined aquifers. In contrast, the LCRB geologic setting features deep alluvial aquifers comprised of inter-bedded sands and silts with varying connectivity [ 4 , 5 , 6 , 7 ]. These distinct structures generate highly non-uniform rates of recharge, groundwater-surface water interactions, and aquifer reactions to stress, which are increasingly being tested by a historic megadrought [ 8 , 9 , 10 , 11 , 12 , 13 , 14 ]. The most extreme drought in 1,200 years accelerates hydrological imbalances across the CRB [ 15 , 16 , 17 , 18 ]. Declining precipitation, snowpack, reservoir storage, and rising atmospheric demand have significantly reduced natural recharge in the basin [ 19 , 20 ]. On the other hand, groundwater is increasingly being used to compensate for shortfalls in surface water, especially in central and southern Arizona [ 21 ]. The potential reduction of Colorado River allocations under post-2026 guidelines only heightens the urgency for groundwater resilience, particularly in central and southern Arizona [ 22 ]. Central and Southern Arizona - particularly the subbasins in and around the Active Management Areas (AMAs) -have a unique place in the LCRB where regional hydrological pressures combine with diverse geological and water use patterns [ 23 ]. This region is in the Basin and Range Province, an arid region that contains deep, unconsolidated alluvial aquifers in fault-bounded basins with large amounts of water stored (see section 5.1). Still, accessibility varies throughout the region [ 24 , 25 ]. Recharge in these aquifers relies on episodic infiltration in mountain fronts and along flowing river channels, resulting in heterogeneous responses to natural and anthropogenic drivers [ 26 , 27 ]. Groundwater was historically a primary water source in Central Arizona until the Central Arizona Project (CAP) was built in the 1980s, allowing for large-scale imports of Colorado River Water [ 28 ]. Rapid urban growth, multi-year drought, and surface water cutbacks have recently renewed reliance on groundwater in the region [ 29 , 30 , 27 , 22 ]. While regulation and Managed Aquifer Recharge (MAR) programs like the Arizona Water Banking Authority (AWBA) have improved aquifer depletion, groundwater declines continue—particularly outside of regulated areas—because of pumping overdraft, limited natural recharge, and variable spatially specific management [ 30 ]. This highlights the importance of diagnosing assessments that incorporate surface climate variability, aquifer response variability, and human management signals to determine vulnerabilities and inform regional groundwater planning with a view to sustainability [ 21 ]. Satellite gravimetry has dramatically increased the potential for large-scale groundwater monitoring [ 31 , 32 , 33 ]. NASA's Gravity Recovery and Climate Experiment (GRACE) and its successor GRACE-Follow On (GRACE/FO) (to refer to the combined GRACE and GRACE-FO from here on) have allowed regular observation of changes in total Terrestrial Water Storage (TWS) since they launched in 2002, through the measurement of tiny shifts in Earth's gravity field [ 34 , 35 ]. Groundwater Storage Anomalies (GWSA) are calculated by removing observations or model simulations of snow, soil moisture, and surface water from TWS, providing monthly measures of downward integrated groundwater change rates [ 21 , 36 , 37 , 38 , 39 ]. The global coverage and consistency of GRACE/FO are particularly advantageous in regions where very few in situ monitoring sites exist [ 40 ]. However, although GRACE/FO provides reliable measures of water storage changes, it cannot identify the surface processes by which that change occurs [ 41 ]. Several recent studies indicate that combining GRACE/FO with land surface models improves the capacity to diagnose groundwater trends and climate drivers related to those trends [ 42 , 43 , 44 ]. Few of these studies have facilitated actionable insights for practitioners in planning contexts. Researchers have increasingly turned to land surface models like NASA's Western Land Data Assimilation System (WLDAS) [ 45 ]. This simulates physical processes across the land-atmosphere interface [ 46 ]. WLDAS is built upon the Noah-MP Land Surface Model [ 47 ] and incorporates satellite and in-situ data to simulate surface fluxes such as evapotranspiration, runoff, and snowmelt, and storages such as soil moisture, with high spatial and temporal scales [ 48 ]. WLDAS offers a consistent and high-resolution view of surface conditions influencing groundwater recharge, especially in data-scarce and topographically diverse regions like Arizona. We present a diagnostic approach integrating satellite-based groundwater observations with land surface model data to assess subbasin-scale hydroclimatic controls on groundwater storage. We test our approach in central and southern Arizona region. By combining GRACE/FO-derived GWSA with WLDAS hydrometeorological variables, this study captures both the spatial variability and temporal dynamics of surface–subsurface interactions across Central and Southern Arizona. Our approach centers on three core steps: (1) quantifying long-term trends in groundwater and surface conditions using standardized anomalies, (2) identifying key climate–groundwater linkages through correlation analysis, and (3) extracting dominant hydroclimatic modes via Principal Component Analysis (PCA). We further classify subbasins into similar groups using clustering of PCA scores to enhance interpretability and support spatial management planning. This multi-method approach enables a robust assessment of how groundwater variability responds to climate and offers practical insight into regional groundwater vulnerability under future water stress scenarios. 2. Results 2.1. Spatial patterns of long-term trends The spatial distributions of standardized linear trends in GRACE/FO-derived GWSA and WLDAS surface variables from 2004 to 2021 (Fig. 1 ) show pronounced hydroclimatic heterogeneity in the study area. These trends reflect the combined effects of atmospheric forcing, surface water fluxes, and storage processes, indicating distinct climate–water trajectories across subbasins. The southern subbasins show increased surface temperature, air temperature, total evapotranspiration, and vegetation transpiration, signaling a sustained increase in atmospheric water demand and land surface energy flux. This also corresponds with declines in deep soil moisture. The simultaneous increases in evapotranspiration and decreases in all the storage variables lead to a net groundwater loss trajectory - in line with trends observed in other arid and semi-arid basins experiencing climate-driven groundwater decline [ 22 , 49 ]. In comparison, the northern and central subbasins show a positive trend in both rainfall rate and deep soil moisture, with limited increases in surface and subsurface runoff. These hydrologic anomalies suggest increased infiltration and possible recharge. In these basins, the GRACE/FO GWSA were generally neutral to mildly positive, suggesting the combination of improved moisture inputs and the regional Managed Aquifer Recharge (MAR) programs in some areas, such as those operated by Central Arizona Groundwater Replenishment District (CAGRD) [ 22 , 29 ]. Runoff patterns exhibit substantial spatial variability. For surface runoff, positive trends are confined to the northern and part of the central basins with increasing precipitation and humidity. The trends in subsurface runoff show a similar pattern but are less consistent; nevertheless, there are significant increases in parts of the north and isolated parts of the central basins. 2.2. Correlation of GRACE/FO GWSA and surface fluxes Figure 2 shows how GRACE/FO-derived GWSA and selected WLDAS hydroclimatic variables correlate at the subbasin level. This provides a diagnostic view of the relationship between natural surface–atmosphere processes and groundwater variability throughout the study area. The positive correlations between GRACE/FO GWSA and precipitation rate, deep soil moisture, and subsurface runoff are most pronounced in places in the northern and some central subbasins. In these areas, storage changes are closely related to hydrologic inputs and provide evidence for recharge-dominated processes, in which seasonal to interannual climate variability directly impacts subsurface water storage. This correlation is consistent with conditions in relatively undeveloped basins or those with managed recharge facilities that enhance infiltration and deep percolation [ 50 , 51 ]. On the other hand, southern subbasins show high negative standardized correlations between GRACE/FO GWSA and evapotranspiration, vegetation transpiration, and surface runoff. These subbasins are an example of loss-driven regimes, with strong atmospheric water demand, large land surface energy flux, and limited infiltration capacity; the inverse relationship between storage and surface fluxes shows vulnerability to increased evaporative stress and potential for continued aridification [ 50 ]. The inverse relationship between storage and surface fluxes reflects vulnerability to evaporative stress and progressive aridification, which are unlikely to reverse without supplemental recharge or reduced extraction [ 52 , 53 ]. Interpreting these correlations requires careful consideration of differences in data composition. GRACE/FO GWSA anomalies represent total terrestrial water storage changes, integrating natural hydroclimatic variability and anthropogenic influences such as pumping, artificial recharge, and aquifer management. WLDAS variables, by contrast, are outputs from a physically based land surface model that represents only the natural water and energy balance, excluding human water use [ 47 ]. This distinction is crucial in AMAs—including Phoenix, Pinal, and parts of Tucson—where artificial recharge from CAP deliveries contributes substantially to storage. In such basins, positive GRACE/FO trends or weak correlations with natural variables may be partly driven by Managed Aquifer Recharge (MAR) or other human interventions rather than climate inputs alone (See supplemental information Figure S7). Additionally, to test against serial correlation, we computed cross-correlations at lags − 3 to + 3 months for the 24 target subbasins and re-tested zero-lag significance using an AR(1) effective degrees-of-freedom adjustment. Median patterns indicate that evapotranspiration and vegetation transpiration tend to lead GWSA by about one to two months; rain rate and surface runoff also show the highest at positive lags (variables leading GWSA by roughly one to two months); deep soil moisture peaks near zero lag; and subsurface runoff remains positive across lags. Because GWSA is strongly persistent, the degree of freedom correction reduces nominal significance, yet a non-trivial subset of basins retains detectable zero-lag coupling after adjustment. These coherent lead–lag structures support our variance-attribution result that natural modes account for about one-sixth of spatial trend differences, with evapotranspiration contributing the largest share. Complete distributions, heatmaps, autocorrelation histograms, and significance summaries are provided in the Supplementary Information text S3 and figures S6 to S8. These findings highlight the importance of integrating human water-use datasets in future analysis to partition the respective contributions of climate forcing and anthropogenic management. Without this integration, correlations between GRACE/FO GWSA and natural variables risk conflating climate–groundwater coupling with infrastructure-driven storage changes, especially in regions where policy and management interventions strongly mediate the hydrologic signal. Because GRACE/FO integrates human and natural signals, whereas WLDAS represents only natural processes, correlation strength should not be interpreted as causation, particularly within AMAs, where MAR and pumping can decorrelate climate-storage coupling. Conversely, anthropogenic withdrawals can dampen or reverse the natural correlation signal in heavily pumped basins without significant artificial recharge. For example, in agricultural zones with high extraction rates and limited water imports, GRACE/FO GWSA may show declining trends even where WLDAS variables indicate favorable natural recharge conditions [ 21 ]. 2.3. Principal Component Analysis (PCA) PC1 loads with similar magnitude and sign across all six variables (see tables S1 and S2), representing a coherent hydro-climate coupling mode. PC2 separates atmospheric flux coupling from runoff/plant-use pathways, while PC3 emphasizes deep-soil storage (SM100–200) opposed by plant use/quick runoff. These patterns are spatially consistent with correlation maps, delineating a recharge-responsive north/central tier and a loss-dominated southern tier, with AMA basins showing management-modulated signals. PCA score maps (Fig. 3 ) show coherent regional groupings across the mapped subbasins (Fig. 5 ). PC1 is low in basins such as the East Salt River Valley (ESR), West Salt River Valley (WSR), and Harquahala (HAR). At the same time, it is high in Aguirre Valley (AGV) and Santa Cruz AMA (SCA), with Avra Valley (AVR) and Maricopa–Stanfield (MST) positive to near neutral. PC2 separates atmospheric-flux versus runoff/plant-use pathways scores that are low/negative in HAR, Gila Bend (GIL), Agua Fria (AGF), and positive in SCA. PC3 emphasizes deep-soil storage versus plant-use/quick-runoff. Scores are near zero in ESR, modestly positive in MST, Eloy (ELO), Vekol Valley (VEK), and Gila Bend (GIL), and higher positive in Santa Rosa Valley (SRO). We also examined how long-term groundwater storage changes relate to identified PCA-derived hydroclimatic modes. Regression of basin trends from GRACE/FO-derived subbasins onto the first three PCs explained about 16% of the spatial variance, which indicates natural variability explained a small but measurable part of the observed differences between subbasins. Among the explained variance, total evapotranspiration had the most significant contribution (≈ 28.6%), followed by rainfall (≈ 23.4%) and subsurface runoff (≈ 19.8%), while vegetation transpiration (≈ 13.0%), surface runoff (≈ 10.0%), and deep soil moisture (≈ 5.2%) had smaller contributions. These values are consistent with Arizona's hydroclimatic setting. High evaporative demand is responsible for a significant share of water loss, precipitation, and mountain-front recharge as essential inputs, and deep infiltration signals are degraded and less pervasive at the basin scale. Across the study area, recharge-related processes accounted for approximately 58% of the explained variance compared with loss processes, which accounted for 42%. Given the modest R², these shares should be viewed as directional rather than definitive. This reflects the broader theme that spatial variability in recharge potential explains why specific subbasins experience slower decline rates than others. For management purposes, this brings attention to the need to protect recharge corridors and/or maximize opportunities for recharge (infiltration) in basins where recharge is favored; however, it acknowledges that losses due to the atmosphere and water use will continue to constrain resilience in hotter and drier subbasins operatively. 3. Discussion This study comprehensively evaluates hydroclimatic controls on groundwater storage variability in Central and Southern Arizona, combining GRACE/GRACE-FO satellite gravimetry observations, WLDAS-modeled surface fluxes, and PCA-based dimensionality reduction. The results offer new insights into the spatial heterogeneity of groundwater–climate interactions and reveal distinct hydroclimatic modes that help explain observed depletion patterns, resilience, and management vulnerability. Principal Component Analysis results (Fig. 3 ) indicate that subbasins do not respond uniformly to climate forcing. PC1 is a hydro-climate coherence mode reflecting the overall strength of common coupling between GWSA and precipitation, soil-moisture, runoff, and ET, not a recharge–loss gradient. The recharge–loss contrast emerges on PC2 and PC3: PC2’s sign distinguishes an atmospheric-flux pathway (higher/positive PC2: ET and P side) from a runoff/plant-use pathway (lower/negative PC2: Qsb, Qs, and Tveg). In contrast, higher PC3 indicates stronger deep-soil-storage control. Higher PC3 (and the PC2 sign) mark more recharge-responsive behavior—often amplified where managed aquifer recharge is present. In contrast, low/negative PC2 and/or low PC3 indicate loss-dominated regimes in which atmospheric demand and plant water use/quick runoff decouple GWSA from recharge. Subbasins in this latter group also show negative GWSA trends and weak or negative correlations with recharge variables, reinforcing their vulnerability to hydrologic stress. One of the key challenges in interpreting GRACE/FO-based groundwater anomalies lies in disentangling natural hydroclimatic variability from anthropogenic influences such as pumping and artificial recharge. While the PCA is based on correlations between GRACE/FO (which includes both human and natural signals) and WLDAS (which models only natural processes), mismatches in certain AMAs—such as Phoenix and Pinal—highlight the limitations of relying on climate data alone. These patterns likely reflect intensive water management, where Central Arizona Project (CAP) imports, recharge basins, and regulatory frameworks mask the true extent of climatic stress. The divergence between GRACE/FO and WLDAS in such regions reinforces the need for dual attribution frameworks that separately analyze natural and anthropogenic controls. Without accounting for managed recharge, models may underestimate the buffering effect of infrastructure. Similarly, excluding pumping data prevents a complete diagnosis of overdraft severity. Future studies could incorporate modeled or observed groundwater extraction datasets or use inversion techniques to estimate human extraction as a residual from the GRACE/FP–WLDAS signal mismatch. Clustering of the subbasin PC scores partitions the region into four hydroclimatic regimes (Fig. 4 ). Cluster 4 (dark green) dominates the southern tier—covering most of the agricultural valleys—and represents loss-dominated, climate-sensitive behavior (low/negative PC3 and low PC2). Cluster 3 (blue) forms a south-central belt, indicating mixed controls where atmospheric demand and limited recharge both matter and where management signals are likely. Cluster 2 (green) is concentrated along the north-central, higher-relief corridor, reflecting more recharge/storage-responsive behavior (higher PC2/PC3). Cluster 1 (light) appears in a few central/western rim basins and represents weaker or transitional coupling. This segmentation is consistent with the well-trend and GRACE patterns—Cluster 4 aligns with widespread declines, Clusters 2–3 include more stable or improving conditions, and provide a management-ready way to target pumping curbs, recharge expansion, and monitoring effort. These findings can inform groundwater governance under Arizona's post-2026 water management landscape. With Colorado River shortages likely to persist, reliance on groundwater will intensify, particularly in agriculture-heavy areas like Pinal AMA [ 21 ]. The PCA and clustering results highlight which subbasins are least capable of absorbing additional stress without crossing sustainability thresholds. A key policy takeaway is the need for spatially differentiated management. Instead of uniform restrictions or recharge targets across AMAs, management strategies are more effective if they reflect each subbasin's hydroclimatic regime and recharge potential. For instance, basins in Cluster 4 may require stricter pumping limits or subsidized artificial recharge, while Cluster 1 basins may benefit more from preserving infiltration zones or incentivizing natural recharge. Subbasin classification based on hydroclimatic drivers reveals clear management priorities across Central and Southern Arizona. Loss-dominated basins are highly climate-sensitive, with low natural recharge and high evaporative demand, requiring long-term conservation and Managed Aquifer Recharge (MAR) support. Mixed-stress basins face moderate recharge potential and intense pumping pressure, highlighting the need for tighter integration of surface water, climate, and groundwater management. Recharge-responsive areas show favorable conditions for infiltration and offer strategic opportunities to expand recharge programs. In contrast, compound-risk basins like Gila Bend are more vulnerable, climate-stressed, and unregulated. Gila Bend exemplifies the risks of unmanaged extraction outside AMA boundaries. These results underscore the need for an adaptive, risk-based groundwater policy addressing natural vulnerability and institutional gaps. The discussion of PCA and clustering results is compared to well-level groundwater trends observed by the Arizona Department of Water Resources (ADWR) (See text S2 and figures S2, S3, and S4). Areas identified as climatically sensitive and loss-dominated correspond with depleted subbasins, indicating long-term groundwater decline. In contrast, basins with high PC1 scores and high recharge correlations show fewer over-drafted wells, supporting their classification as more resilient. In-situ groundwater monitoring using data collected from ADWR index wells (located in unconfined aquifers) shows substantial differences in spatial and management-based depth-to-water trends across central and southern Arizona from 2000 to 2023. Subbasin locations in the south and western areas of the study demonstrate extensive and steep declines (red shading in Figure S3; red wells in Figure S4), reflecting groundwater intensive use and low natural recharge. Conversely, groundwater conditions in some northern and eastern subbasins have shown relatively stable or rising circumstances (blue shading in Figure S4; blue wells in Figure S5), frequently correlated with managed recharge activities and lower groundwater withdrawal pressure. Time-series analysis (Figure S5) confirmed notable, statistically significant managed groundwater rising levels AMAs and declining levels in non-AMAs, confirming groundwater governance plays a role in securing declines in water levels. Nonetheless, declining water levels in parts of the AMAs suggest that application of regulatory protections is insufficient unless protection measures include interventions to address climate-driven pressures and groundwater extraction levels. The comparison between this study’s results and the latest AMA’s Management Plans (4MPs) (available online at: https://www.azwater.gov/fourth-management-plan ) shows that management plans may only align with at least partially the hydroclimatic trends outlined by GRACE/FO and WLDAS. There are identified recharge-resilient basins in the northern and central best-practice basins, and those have associated recharge and Assured Water Supply programs; however, these two are static and do not incorporate information from hydro-climatic variables. Transition basins, such as in the Santa Cruz and Pinal AMAs, use conservation programs to address possible issues, but use a fixed evapotranspiration value that does not address the changing climate stress. The intermediate basins remain stable but have weak evolving protection of recharge corridors or management on recovery siting. The Gila Bend subbasin is a significant case illustrating compounded vulnerability from natural and anthropogenic pressures. Gila Bend shows low PCA scores and high correlations with evaporative demand, indicating a natural hydroclimatic deficit. It also lies outside the boundaries of any AMA, leaving it largely exempt from pumping regulations. Long-term monitoring well data confirms steep groundwater declines, suggesting that this subbasin is exposed to climatic stress and unregulated extraction without large-scale artificial projects or management plans. This combination of natural and anthropogenic factors places the Gila Bend subbasin among the highest-risk basins in the study region. In short, while the 4MPs recognize recharge and conservation, they miss climate variability, ET change, and basin-specific needs, leaving room for more adaptive, science-based sustainability plans. This study offers a practical framework for prioritizing subbasins under the post-2026 water management scenario. Loss-dominated clusters may require stricter pumping regulation, expanded recharge infrastructure, and enhanced monitoring, whereas recharge-responsive clusters may benefit from preserving infiltration zones and sustaining artificial recharge operations. Additionally, the diagnostic approach presented here directly relates to physical and data-driven modeling approaches. In the case of process-based physical models, understanding the balance between natural hydroclimatic drivers and anthropogenic influences is essential to represent recharge, evapotranspiration, and storage change mechanisms accurately. In the Machine Learning (ML) and Artificial Intelligence (AI) models, the PCA structure provides an interpretable feature space that captures dominant patterns of groundwater–climate coupling, reducing the risk of overfitting to noise. Without explicit knowledge of the natural and human processes shaping groundwater variability, physical models risk misrepresentation, and without identifying the leading principal components, ML models risk misattribution of drivers Several sources of uncertainty must be acknowledged within the scope of our analysis. First, the GRACE/FO satellite data, while invaluable for detecting regional groundwater changes, are spatially coarse and reflect total Terrestrial Water Storage (TWS), which includes anthropogenic influences such as pumping and managed aquifer recharge. This aggregation makes it challenging to isolate purely natural signals without additional attribution methods. Second, the WLDAS dataset, while physically consistent and spatially detailed, models only natural processes and omits human water use, infrastructure, and policy interventions—factors that are increasingly dominant in groundwater dynamics across central and southern Arizona. These model design and scope differences can introduce signal mismatches in correlation and PCA analyses, particularly in AMAs where human intervention is substantial. Moreover, the land surface model's performance depends on the choice of meteorological forcing and calibration datasets, which can further influence simulated surface fluxes like evapotranspiration or runoff. To partially address these uncertainties, this study employed GRACE/FO products that assimilate observations and WLDAS outputs validated against streamflow and evapotranspiration. Future work can reduce uncertainty by integrating direct pumping data, MAR volumes, and socio-institutional variables. Such additions would enable a complete dual-attribution framework that distinguishes climate-driven variability from anthropogenic impacts, thereby enhancing the utility of satellite-model integration for groundwater governance. Extending this analysis with dynamical modeling—such as groundwater flow or water balance simulations—would strengthen causal attribution and scenario planning. Applying PCA in moving windows (e.g., 5-year intervals) may reveal shifts in dominant controls as climate stress and policy responses unfold. 4. Conclusion This study presents an integrated assessment of the hydroclimatic controls causing groundwater storage variability across central and southern Arizona through the integration of GRACE/FO satellite gravimetry observations, hydrometeorological products from the Western Land Data Assimilation System (WLDAS), and multivariate statistical methods (e.g., Principal Component Analysis and k-means clustering). The loss-dominated subbasins feature groundwater trends that are consistently negative over time, and a strong relationship with evaporative fluxes and low PCA scores. These basins typically have little natural recharge potential, high atmospheric demand for water, and high sensitivity to warm and dry conditions made worse by climate effects, making them highly prone to groundwater depletion via climate stress. Meanwhile, we found that the recharge-responsive subbasins are highly associated with groundwater storage and natural recharge-related variables (i.e., rainfall, deep soil moisture, runoff), had higher PCA scores, and spatial clustering indicating higher resilience capacities. Extended artificial recharge schemes could enhance beneficial hydroclimatic preconditions in many of these basins. The regime map is diagnostic, not causal. It highlights where climate coupling is strong vs weak; targeted management data (pumping, MAR volumes) are needed to complete attribution. The diagnostic approach developed in this study is transferable to other regions with arid and semi-arid climate stress and changing surface water availability. Future work may incorporate human water-use metrics, artificial recharge volumes, and dynamic groundwater modelling to facilitate causal attribution efforts. PCA moving-window analysis could also discern shifts in subbasin sensitivity over timescales as conditions change and management actions occur. As Arizona and the Lower Colorado River Basin region manage more extended drought periods and changing surface water allocations, future water managers will need tools considering natural hydroclimatic variability and anthropogenic drivers to develop differentially focused, climate-informed, sustainability-based groundwater policies. 5. Methods 5.1. Study Area Central and Southern Arizona are critical to evaluating groundwater vulnerability and sustainability in the Lower Colorado River Basin (LCRB) [ 22 ]. The region contains the Phoenix, Pinal, and Tucson Active Management Areas (AMAs)—bounded jurisdictions established to prevent groundwater overdraft and by the Arizona 1980 Groundwater Management Act [ 24 , 25 ]. The AMAs are located within the ecologically arid to semi-arid Basin & Range Province. They are characterized by large, fault-contained, alluvial basins with accumulations of unconsolidated material—with high potential for large-capacity aquifer systems, more prone to significant aquifer drawdown, subsidence, and degradation in response to an extended duration of natural and anthropogenic stressors [ 54 , 55 , 56 ]. Central Arizona is a microcosm of layered water stress, where urban demand on a finite supply converges with irrigated agriculture, surface water availability diminished by human-caused climate change, and institutional constraints on supply and pumping [ 29 , 57 ]. The surface water supply comes primarily from the Central Arizona Project (CAP) infrastructure, which is a 336-mile canal that delivers the State of Arizona's share of Colorado River water, and, to a lesser extent, local runoff from the Salt and Verde Rivers- not to mention the local micro-climate influences on the available surface water supply. However, CAP allocations are becoming more uncertain due to the prolonged megadrought, basin-wide over-allocation, and declining snowpack and runoff in the UCRB [ 2 , 58 ]. Agricultural users are already increasingly dependent on groundwater for supply gaps, especially in places where CAP reductions have forced crop fallowing and re-drilling wells ever deeper [ 59 ]. Recharge in these basins is variable, typically low, precipitation-driven, streamflow-driven, and mountain-front driven [ 60 ]. Natural recharge rates are mostly low and spatially variable. However, managed recharge through basins and direct injections coordinated by the Arizona Water Banking Authority (AWBA) has become an essential means of groundwater recovery in AMAs where water may be available from CAP [ 61 , 62 ]. Not all subbasins benefit in equal degrees from such institutionally structured interventions; for example, regions of central Arizona have experienced over 100 meters of groundwater level decline and six meters of subsidence [ 63 ]. Therefore, focusing on Central and Southern Arizona presents an interesting testbed for scalable groundwater governance mechanisms, especially over the uncertain surface water demand and increasing frequency and intensity of climate extremes. 5.2. Analytical workflow This study employs a multi-step analytical workflow to quantify the influence of main hydroclimatic drivers on groundwater storage variability across groundwater subbasins in central and southern Arizona. The analysis consists of three significant steps: (1) anomaly calculation, standardization, and linear trend calculations, (2) correlation analysis, and (3) dimensionality reduction and clustering to identify the dominant hydroclimatic mode (Fig. 6 ). For full methodological details, including equations and notation, see Supplementary Text S1–S3. All the time series—including GRACE/FO-derived groundwater storage anomalies (GWSA) and land surface variables from WLDAS—were converted to standardized anomalies in the first stage. We formed standardized anomalies by subtracting the monthly climatology and dividing by the interannual standard deviation. Linear trends were then estimated on these standardized anomalies. Linear trend magnitudes were estimated via ordinary least squares regression at each 0.25° spatial grid cell [ 64 ], and slopes were subsequently standardized across the domain to allow spatial comparison of long-term changes in surface and subsurface water conditions. Next, the Pearson correlation coefficients were computed between standardized GRACE/FO anomalies and each WLDAS variable at the grid level [ 63 ]. These correlations were aggregated to the subbasin scale using zonal means, producing a subbasin-by-variable matrix that summarizes the strength and direction of hydroclimatic linkages across the study area. PCA was applied to the correlation matrix to extract the dominant modes of spatial variability [ 64 ]. The leading components, PC1 and PC2, identify the primary surface processes (e.g., evapotranspiration, runoff, soil moisture) influencing subbasin-scale groundwater variability. Subbasins were embedded in a two-dimensional feature space defined by their PC1 and PC2 scores to identify spatial patterns in hydroclimatic control. K-means clustering [ 65 ] was then applied to classify subbasins into four groups with similar hydroclimatic signatures. The scikit-learn implementation of the K-Means algorithm was used with a fixed random seed to ensure reproducibility [ 66 ]. For further comparison, long-term trends in groundwater depth were estimated using in-situ ADWR Index Wells data across selected subbasins (see Supplementary Text S3 and Figures S4 and S5). 5.3. Data This study integrates satellite observations and model-derived datasets to examine the hydroclimatic controls on groundwater storage variability across central and southern Arizona. Specifically, we utilize (1) groundwater storage anomalies from the GRACE/FO missions, which are extracted from observations of Total Water Storage (TWS) [ 49 ], (2) land surface hydrometeorological variables from the Western Land Data Assimilation System (WLDAS) [ 45 ]. These datasets collectively enable a multiscale assessment of groundwater and climate interactions from 2004 to 2021. Table 1 and Figure S2 describe the datasets used in this study. Table 1 Summary of datasets Data Source Variables Spatial Resolution Temporal Span GRACE/FO Groundwater Storage Anomalies (bias-corrected, enhanced resolution) • Jet Propulsion Laboratory (JPL) • Chandanpurkar et al., 2025 • Total Water Storage Anomaly (TWSA) • Groundwater Storage Anomaly (GWSA) 0.25° × 0.25° 2004–2021 WLDAS Land Surface Variables Western Land Data Assimilation System (WLDAS) (Erlingis et al., 2021) • Surface Temperature (Avgsurft_Tavg), • Air Temperature (Tair_F_Tavg), • Specific Humidity (Qair_F_Tavg), • Rain Precipitation Rate (Rainf_Tavg), • Surface Runoff (Qs_Tavg), • Subsurface Runoff (Qsb_Tavg), • Total Evapotranspiration (Evap_Tavg), • Vegetation Transpiration (Tveg_Tavg), • Soil Moisture at 100–200 Cm Depth (Soilmoi100_200cm_Tavg), • Groundwater Storage (GWS_Tavg) 0.01° × 0.01° 2004–2021 Table 2 Classification of central and southern Arizona subbasins. Cluster Hydroclimatic interpretation Management implications Cluster 1 – Recharge-responsive Intermediate recharge signals, relatively neutral stress. PCA shows moderate positive coupling with soil moisture and precipitation, weak ET influence. Transitional case: could respond positively to Managed Aquifer Recharge (MAR) expansion; needs monitoring to ensure pumping doesn’t tip balance toward depletion. Cluster 2 – Recharge-dominated High positive correlations of GWSA with precipitation, deep soil moisture, and subsurface runoff. Weak loss signals. PCA scores show high recharge coupling. Natural recharge potential is relatively high. Protect infiltration corridors (mountain fronts, ephemeral washes); MAR complements high natural signals. Cluster 3 – Transition basins Mixed recharge–loss signatures; moderate soil-moisture/runoff coupling but elevated ET/transpiration stress. PCA scores intermediate. Many coincide with AMA urban/ag centers. Outcomes hinge on management. climate stress. Long-term vulnerability if pumping >(natural + managed recharge). Proactive management plans are necessary. Cluster 4 – Loss-dominated, high stress Weak recharge signals; high negative correlations of GWSA with evapotranspiration and vegetation demand. PCA shows low/negative PC2–PC3. Highest compound risk: climate stress + weak regulation. Persistent Depth To Water (DTW) declines confirm loss-dominated status. Require stricter overdraft controls. MAR feasibility may be limited due to poor natural recharge capacity. WLDAS provides high-resolution, observation-driven estimates of key land surface hydrologic variables such as precipitation, evapotranspiration, soil moisture, and runoff. Developed to support drought monitoring and water resources research in the western United States. WLDAS integrates multiple remote sensing and in situ data sources with advanced land surface models to produce gridded outputs at 0.01° spatial resolution. While WLDAS offers valuable insights into land-atmosphere processes and captures hydroclimatic variability at fine spatial scales, uncertainties remain due to model structural limitations, errors in meteorological forcing inputs, and challenges in representing subsurface hydrology and anthropogenic water use. Despite these limitations, WLDAS datasets have been widely used for regional water balance studies and serve as a critical complement to satellite-based storage observations in this study [ 67 , 68 , 69 ]. The GRACE/FO TWS data are based on the Landerer Mascon [ 34 ] Release 6 Version 3 solution, which offers improved spatial resolution (0.25° × 0.25°) through post-processing that integrates Global Land Data Assimilation System (GLDAS) models [ 70 ] and corrects for biases at the mascon scale [ 49 ]. This enhancement enables meaningful sub-mascon regional analysis while retaining consistency with the original GRACE observations. The processed dataset accounts for significant sources of uncertainty by incorporating JPL-provided formal errors and climatology-removed anomaly estimation. To estimate groundwater storage changes from GRACE/FO TWS, non-groundwater signals (e.g., snow water equivalent, surface and soil water storage) must be calculated and removed [ 38 ]. This study utilizes global enhanced-resolution, bias-corrected groundwater storage anomaly data as processed and described by [ 49 ]. In our study, these data support the investigation of groundwater storage variability across central and southern Arizona, where hydrologic and climatic conditions require high-resolution satellite-informed datasets to assess groundwater–climate interactions over the 2004–2021 period. Declarations Competing interests The authors declare no competing interests. Additional information Supplementary Information Funding This research was supported by funding from the Arizona State University School of Sustainability and College of Global Futures. Author Contribution J.S.F. secured funding, provided resources, and supervised the project. B.M. and J.S.F. designed and conceptualized the study. B.M. performed the formal analysis, developed the software, carried out validation, and prepared the visualizations. B.M. and J.S.F. contributed to the investigation and methodology. J.S.F. led project administration. B.M. and J.S.F. wrote the original draft. All authors contributed to reviewing and editing the manuscript. Data Availability Monthly hydrometeorological data were obtained from the NASA Western Land Data Assimilation System (WLDAS), specifically the Noah-MP Land Surface Model L4 daily product at 0.01° resolution ( [https://hydro1.gesdisc.eosdis.nasa.gov/data/WLDAS/WLDAS_NOAHMP001_DA1.D1.0/](https:/hydro1.gesdisc.eosdis.nasa.gov/data/WLDAS/WLDAS_NOAHMP001_DA1.D1.0) ) The data were accessed on November 4, 2024, via NASA's Earthdata system using authenticated download links. A bash automated script was used to retrieve and organize the NetCDF files locally for analysis. GRACE/GRACE-FO Mascon data are available at [https://podaac.jpl.nasa.gov/dataset/TELLUS_GRAC-GRFO_MASCON_CRI_GRID_RL06.3_V4](https:/podaac.jpl.nasa.gov/dataset/TELLUS_GRAC-GRFO_MASCON_CRI_GRID_RL06.3_V4) . References US Bureau of Reclamation. (2012). Colorado River Basin Water Supply and Demand Study. https://www.usbr.gov/lc/region/programs/crbstudy/finalreport/Study%20Report/CRBS_Study_Report_FINAL.pdf Richter, B. D., Lamsal, G., Marston, L., Dhakal, S., Sangha, L. S., Rushforth, R. R., Wei, D., Ruddell, B. L., Davis, K. F., Hernandez-Cruz, A., Sandoval-Solis, S., & Schmidt, J. C. (2024). New water accounting reveals why the Colorado River no longer reaches the sea. Communications Earth and Environment, 5(1). https://doi.org/10.1038/s43247-024-01291-0 Fleck, J., & Udall, B. (2021). Managing Colorado River risk. Science, 372(6545), 885. https://doi.org/10.1126/SCIENCE.ABJ5498 Hunt, C., & Rabbitt, M. (1969). Geologic history of the Colorado River. In the US Geological Survey. https://pubs.usgs.gov/pp/0669/report.pdf#page=75 Guay Jacobs, K. L., & Holway, J. M. (2004). Managing for sustainability in an arid climate: Lessons learned from 20 years of groundwater management in Arizona, USA. Hydrogeology Journal, 12(1), 52–65. https://doi.org/10.1007/s10040-003-0308-y Howard, K., & Bohannon, R. (2001). Lower Colorado River: Framework, neogene deposits, incision, and evolution. https://pubs.usgs.gov/publication/70198241 Crossey, L. C., Karlstrom, K. E., Dorsey, R., Pearce, J., Wan, E., Beard, L. S., Asmerom, Y., Polyak, V., Crow, R. S., Cohen, A., Bright, J., & Pecha, M. E. (2015). Importance of groundwater in propagating downward integration of the 6–5 Ma Colorado River system: Geochemistry of springs, travertines, and lacustrine carbonates. Geosphere , 11, 660–682. https://doi.org/10.1130/GES01073.1 Woodhouse, C. A., Pederson, G. T., Morino, K., McAfee, S. A., & McCabe, G. J. (2016). Increasing influence of air temperature on upper Colorado River streamflow. Geophysical Research Letters, 43(5), 2174–2181. https://doi.org/10.1002/2015GL067613 Solander, K. C., Bennett, K. E., & Middleton, R. S. (2017). Shifts in historical streamflow extremes in the Colorado River Basin. Journal of Hydrology: Regional Studies, 12, 363–377. https://doi.org/10.1016/j.ejrh.2017.05.004 Xiao, M., Udall, B., & Lettenmaier, D. P. (2018). On the Causes of Declining Colorado River Streamflows. Water Resources Research, 54(9), 6739–6756. https://doi.org/10.1029/2018WR023153 Tran, H., Zhang, J., Cohard, J. M., Condon, L. E., & Maxwell, R. M. (2020). Simulating Groundwater-Streamflow Connections in the Upper Colorado River Basin. Groundwater, 58(3), 392–405. https://doi.org/10.1111/gwat.13000 Bennett, K. E., Talsma, C., & Boero, R. (2021). Concurrent changes in extreme hydroclimate events in the Colorado River Basin. Water (Switzerland), 13(7). https://doi.org/10.3390/w13070978 Schmidt, J. C., Yackulic, C. B., & Kuhn, E. (2023). The Colorado River water crisis: Its origin and the future. Wiley Interdisciplinary Reviews: Water, 10(6), e1672. https://doi.org/10.1002/WAT2.1672 Hoerling, M. P., Eischeid, J. K., Diaz, H. F., Rajagopolan, B., & Kuhn, E. (2024). Critical Effects of Precipitation on Future Colorado River Flow. Journal of Climate. https://doi.org/10.1175/JCLI-D-23-0617.1 Martin, J. T., Pederson, G. T., Woodhouse, C. A., Cook, E. R., McCabe, G. J., Anchukaitis, K. J., ... & King, J. (2020). Increased drought severity tracks warming in the United States’ largest river basin. Proceedings of the National Academy of Sciences, 117(21), 11328-11336. Gangopadhyay, S., Woodhouse, C. A., McCabe, G. J., Routson, C. C., & Meko, D. M. (2022). Tree Rings Reveal Unmatched 2nd Century Drought in the Colorado River Basin. Geophysical Research Letters, 49(11). https://doi.org/10.1029/2022GL098781 McCabe, G. J., Anchukaitis, K. J., Wise, E. K., Erger, P. J., Dolan, L., McGuire, M., Gangopadhyay, S., Chase, K. J., Littell, J. S., Gray, S. T., George, S. S., Friedman, J. M., Sauchyn, D. J., St-Jacques, J. M., & King, J. (2020). Increased drought severity tracks warming in the United States' largest river basin. Proceedings of the National Academy of Sciences of the United States of America, 117(21). https://doi.org/10.1073/pnas.1916208117 Pokharel, B., Jagannathan, K. A., Wang, S. ‐Y. S., Jones, A., LaPlante, M. D., Buddhavarapu, S., Borhara, K., Ulrich, P., Leung, L. R., Eklund, J., Hasenyager, C., Serago, J., Prairie, J. R., Kaatz, L., Winchell, T., & Kugel, F. (2024). Can we rely on drought‐ending "miracles" in the Colorado River Basin? JAWRA Journal of the American Water Resources Association. https://doi.org/10.1111/1752-1688.13204 Tillman, F. D., Gangopadhyay, S., & Pruitt, T. (2016). Changes in groundwater recharge under projected climate in the upper Colorado River basin. Geophysical Research Letters, 43(13), 6968–6974. https://doi.org/10.1002/2016GL069714 Ryu, J., Wang, S.-Y., Jeong, J.-H., Kim, H., & Yoon, J.-H. (2025). Time of emergence (TOE) of potential aridification in the western United States. Journal of Hydrology, 626, 133029. https://doi.org/10.1016/j.jhydrol.2025.133029 Abdelmohsen, K., Famiglietti, J. S., Ao, Y. Z., Mohajer, B., & Chandanpurkar, H. A. (2025). Declining Freshwater Availability in the Colorado River Basin Threatens Sustainability of Its Critical Groundwater Supplies. Geophysical Research Letters, 52(10). https://doi.org/10.1029/2025GL115593 Scanlon, B. R., Pool, D. R., Rateb, A., Conway, B., Sorensen, K., Udall, B., & Reedy, R. C. (2025). Multidecadal drought impacts on the Lower Colorado Basin with implications for future management. Communications Earth and Environment, 6(1). https://doi.org/10.1038/s43247-025-02149-9 Perry, D., Swanson, R. K., & Springer, A. E. (2024). Policy deficiencies and contingency plans: Groundwater management implications for baseflow contributions to the Colorado River. Frontiers in Environmental Science, 12, 1444015. https://doi.org/10.3389/fenvs.2024.1444015 Tillman, F. D., & Leake, S. A. (2010). Trends in groundwater levels in wells in the active management areas of Arizona, USA. Hydrogeology Journal, 18(7), 1515–1524. https://doi.org/10.1007/s10040-010-0603-3 Jacobs, K. L., & Holway, J. M. (2004). Managing for sustainability in an arid climate: lessons learned from 20 years of groundwater management in Arizona, USA. Hydrogeology Journal, 12(1), 52-65. Garner, B. D., Pool, D. R., Tillman, F. D., & Forbes, B. T. (2013). Human effects on the hydrologic system of the Verde Valley, central Arizona, 1910–2005 and 2005–2110, using a regional groundwater flow model (Scientific Investigations Report 2013–5029). US Geological Survey. https://doi.org/10.3133/sir20135029 Tillman, F. D., & Flynn, M. E. (2024). Arizona Groundwater Explorer: Interactive maps for evaluating the historical and current groundwater conditions in wells in Arizona, USA. Hydrogeology Journal, 32, 645–661. https://doi.org/10.1007/s10040-023-02748-w Hanemann, W. M. (2002). The Central Arizona Project (Working Paper No. 937). Department of Agricultural and Resource Economics and Policy, Division of Agricultural and Natural Resources, University of California at Berkeley. Giannini Foundation of Agricultural Economics. https://escholarship.org/uc/item/9pd083d5 Milman, A., Bonnell, C., Maguire, R., Sorensen, K., & Blomquist, W. (2021). Groundwater Recharge for Water Security: The Arizona Water Bank, Arizona. Case Studies in the Environment, 5(1). https://doi.org/10.1525/cse.2020.1113999 Wahal, A., Mendenhall, E., & Giordano, M. (2023). Water in the West: Analyzing the disconnect between farmers' and policymakers' perceptions of Colorado River Basin shortages in Arizona. Environmental Science & Policy, 146, 57–66. https://doi.org/10.1016/j.envsci.2023.05.002 Rodell, M., Velicogna, I., & Famiglietti, J. S. (2009). Satellite-based estimates of groundwater depletion in India. Nature, 460(7258), 999-1002. Famiglietti, J. S., Lo, M., Ho, S. L., Bethune, J., Anderson, K. J., Syed, T. H., ... & Rodell, M. (2011). Satellites measure recent rates of groundwater depletion in California's Central Valley. Geophysical Research Letters, 38(3). Castle, S. L., Thomas, B. F., Reager, J. T., Rodell, M., Swenson, S. C., & Famiglietti, J. S. (2014). Groundwater depletion during drought threatens future water security of the Colorado River Basin. Geophysical research letters, 41(16), 5904-5911. Tapley, B. D., Watkins, M. M., Flechtner, F., Reigber, C., Bettadpur, S., Rodell, M., Richter Sasgen, I., Famiglietti, J. S., Landerer, F. W., Chambers, D. P., Reager, J. T., Gardner, A. S., Save, H., Ivins, E. R., Swenson, S. C., Boening, C., Dahle, C., Wiese, D. N., Dobslaw, H., … Velicogna, I. (2019). Contributions of GRACE to understanding climate change. In Nature Climate Change (Vol. 9, Issue 5, pp. 358–369). Nature Publishing Group. https://doi.org/10.1038/s41558-019-0456-2 Rodell, M., & Reager, J. T. (2023). Water cycle science enabled by the GRACE and GRACE-FO satellite missions. Nature Water, 1(1), 47–59. https://doi.org/10.1038/s44221-022-00005-0 Liu, P. W., Famiglietti, J. S., Purdy, A. J., Adams, K. H., McEvoy, A. L., Reager, J. T., ... & Rodell, M. (2022). Groundwater depletion in California’s Central Valley accelerates during megadrought. Nature Communications, 13(1), 7825. Rodell, M., Chen, J., Kato, H., Famiglietti, J. S., Nigro, J., & Wilson, C. R. (2007). Estimating groundwater storage changes in the Mississippi River basin (USA) using GRACE. Hydrogeology Journal, 15(1), 159–166. https://doi.org/10.1007/s10040-006-0103-7 Rodell, M., & Famiglietti, J. S. (2002). The potential for satellite-based monitoring of groundwater storage changes using GRACE: the High Plains aquifer, Central US. Journal of Hydrology. www.elsevier.com/locate/jhydrol Yeh, P. J. F., Swenson, S. C., Famiglietti, J. S., & Rodell, M. (2006). Remote sensing of groundwater storage changes in Illinois using the Gravity Recovery and Climate Experiment (GRACE). Water Resources Research, 42(12). Adams, K. H., Reager, J. T., Rosen, P., Wiese, D. N., Farr, T. G., Rao, S., Haines, B. J., Argus, D. F., Liu, Z., Smith, R., Famiglietti, J. S., & Rodell, M. (2022). Remote Sensing of Groundwater: Current Capabilities and Future Directions. In Water Resources Research (Vol. 58, Issue 10). John Wiley and Sons Inc. https://doi.org/10.1029/2022WR032219 Reager, J. T., & Famiglietti, J. S. (2013). Characteristic mega-basin water storage behavior using GRACE. Water Resources Research, 49(6), 3314–3329. https://doi.org/10.1002/wrcr.20264 Ali, S., Liu, D., Fu, Q., Cheema, M. J. M., Pham, Q. B., Rahaman, M. M., Dang, T. D., & Anh, D. T. (2021). Improving the resolution of grace data for spatio-temporal groundwater storage assessment. Remote Sensing, 13. https://doi.org/10.3390/rs13173513 Humphrey, V., Gudmundsson, L., & Seneviratne, S. I. (2016). Assessing Global Water Storage Variability from GRACE: Trends, Seasonal Cycle, Subseasonal Anomalies and Extremes. In Surveys in Geophysics (Vol. 37, pp. 357–395). Springer Netherlands. https://doi.org/10.1007/s10712-016-9367-1 Yin, W., Hu, L., Zhang, M., Wang, J., & Han, S. C. (2018). Statistical Downscaling of GRACE-Derived Groundwater Storage Using ET Data in the North China Plain. Journal of Geophysical Research: Atmospheres, 123(11), 5973–5987. https://doi.org/10.1029/2017JD027468 Erlingis, J. M., Rodell, M., Peters-Lidard, C. D., Li, B., Kumar, S. V., Famiglietti, J. S., Granger, S. L., Hurley, J. V., Liu, P. W., & Mocko, D. M. (2021). A High-Resolution Land Data Assimilation System Optimized for the Western United States. Journal of the American Water Resources Association, 57, 692–710. https://doi.org/10.1111/1752-1688.12910 Billah, M. M., Goodall, J. L., Narayan, U., Reager, J. T., Lakshmi, V., & Famiglietti, J. S. (2015). A methodology for evaluating evapotranspiration estimates at the watershed-scale using GRACE. Journal of Hydrology, 523, 574–586. https://doi.org/10.1016/j.jhydrol.2015.01.066 Niu, G.-Y., Yang, Z.-L., Mitchell, K. E., Chen, F., Ek, M. B., Barlage, M., et al. (2011). The community Noah land surface model with multiparameterization options (Noah-MP): 1. Model description and evaluation with local-scale measurements. Journal of Geophysical Research, 116(D12). https://doi.org/10.1029/2010JD015139 He, C., Valayamkunnath, P., Barlage, M., Chen, F., Gochis, D., Cabell, R., Schneider, T., Rasmussen, R., Niu, G. Y., Yang, Z. L., Niyogi, D., & Ek, M. (2023). Modernizing the open-source community Noah with multi-parameterization options (Noah-MP) land surface model (version 5.0) with enhanced modularity, interoperability, and applicability. Geoscientific Model Development, 16(17), 5131–5151. https://doi.org/10.5194/gmd-16-5131-2023 Chandanpurkar, H. A., Famiglietti, J. S., Gopalan, K., Wiese, D. N., Wada, Y., Kakinuma, K., Reager, J. T., & Zhang, F. (2025). Unprecedented continental drying, shrinking freshwater availability, and increasing land contributions to sea level rise. Science Advances, 11(30). https://doi.org/10.1126/sciadv.adx0298 Thomas, B. F., & Famiglietti, J. S. (2019). Identifying climate-induced groundwater depletion in GRACE observations. Sci Rep 9: 4124. https://doi.org/10.1038/s41598-019-40155-y Liu, P. W., Famiglietti, J. S., Purdy, A. J., Adams, K. H., McEvoy, A. L., Reager, J. T., ... & Rodell, M. (2022). Groundwater depletion in California’s Central Valley accelerates during megadrought. Nature Communications, 13(1), 7825. https://doi.org/10.1038/s41467-022-35582-x. Condon, L. E., Atchley, A. L., & Maxwell, R. M. (2020). Evapotranspiration depletes groundwater under warming over the contiguous United States. Nature Communications, 11, 873. https://doi.org/10.1038/s41467-020-14688-0 Wu, X., Feng, X., & Fu, B. (2025). Climate aridification and intensified human interference undermined water storage in the Lower Yellow River region. Geography and Sustainability, 6(4), 100303. https://doi.org/10.1016/j.geosus.2025.100303 Fred Tillman, B. D., Gangopadhyay, S., & Pruitt, T. (2020). Trends in recent historical and projected climate data for the Colorado River Basin and potential effects on groundwater availability. https://pubs.usgs.gov/publication/sir20205107 Kyl Center for Water Policy at ASU's Morrison Institute. (2024). Groundwater Protection in the Valley of the Sun. https://morrisoninstitute.asu.edu/sites/default/files/2024-04/ASU-KC-GrndwtrRprt-FNL-HR-4.pdf Ohenhen, L. O., Zhai, G., Lucy, J., Werth, S., Carlson, G., Khorrami, M., Onyike, F., Sadhasivam, N., Tiwari, A., Ghobadi-Far, K., Sherpa, S. F., Lee, J.-C., Zehsaz, S., & Shirzaei, M. (2025). Land subsidence risk to infrastructure in US metropolises. Nature Cities, 2, 543–554. https://doi.org/10.1038/s44284-025-00240-y Tadych, D. E., Ford, M., Colby, B. G., & Condon, L. E. (2024). Historical patterns of well drilling and groundwater depth in Arizona, considering groundwater regulation and surface water access. Journal of the American Water Resources Association, 60(6), 1193–1208. https://doi.org/10.1111/1752-1688.13234 Hoerling, M. P., Eischeid, J. K., Diaz, H. F., Rajagopolan, B., & Kuhn, E. (2024). Critical effects of precipitation on future Colorado River flow. Journal of Climate, 37(16), 4079-4093. Arizona Department of Water Resources. (2022, September 6). ADWR's review of Pinal stakeholder model revisions. https://www.azwater.gov/news/articles/2022-09-06/adwrs-review-pinal-stakeholder-model-revisions Qiu, Y., Famiglietti, J. S., Behrangi, A., Farmani, M. A., Sohi, H. Y., Gupta, A., Hung, F., Abdelmohsen, K., & Niu, G.-Y. (2025). The strong impact of precipitation intensity on groundwater recharge and terrestrial water storage change in Arizona, a typical dryland. Geophysical Research Letters, 52(14), e2025GL114747. https://doi.org/10.1029/2025GL114747 Milman, A., Bonnell, C., Maguire, R., Sorensen, K., & Blomquist, W. (2021). Groundwater recharge for water security: the Arizona Water Bank, Arizona. Case Studies in the Environment, 5(1), 1113999. Bernat, R. F. A., Megdal, S. B., Eden, S., & Bakkensen, L. A. (2023). Stakeholder opinions on the issues of the Central Arizona Groundwater Replenishment District and policy alternatives. Water, 15(6), 1166. https://doi.org/10.3390/w15061166 Conway, B. D. (2015). Land subsidence and earth fissures in south-central and southern Arizona, USA. Hydrogeology Journal, 23(6), 1247–1266. https://doi.org/10.1007/s10040-015-1329-z Draper, N. R., & Smith, H. (1998). Applied regression analysis (3rd ed.). Wiley. Rodgers, J. L., & Nicewander, W. A. (1988). Thirteen ways to look at the correlation coefficient. The American Statistician, 42(1), 59–66. https://doi.org/10.1080/00031305.1988.10475524 Abdi, H., & Williams, L. J. (2010). Principal component analysis. In Wiley Interdisciplinary Reviews: Computational Statistics (Vol. 2, Issue 4, pp. 433–459). https://doi.org/10.1002/wics.101 Arthur, D., & Vassilvitskii, S. (2007). k-means++: The advantages of careful seeding. In Proceedings of the eighteenth annual ACM-SIAM symposium on Discrete algorithms (pp. 1027–1035). Society for Industrial and Applied Mathematics. Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., Vanderplas, J., Passos, A., Cournapeau, D., Brucher, M., Perrot, M., & Duchesnay, É. (2011). Scikit-learn: Machine learning in Python. Journal of Machine Learning Research, 12, 2825–2830. Kumar, S. V., Peters-Lidard, C. D., Tian, Y., Houser, P. R., Geiger, J., Olden, S., Lighty, L., Eastman, J. L., Doty, B., Dirmeyer, P., Adams, J., Mitchell, K., Wood, E. F., & Sheffield, J. (2006). Land Information System—An interoperable framework for high resolution land surface modeling. Environmental Modelling & Software, 21(10), 1402–1415. https://doi.org/10.1016/j.envsoft.2005.07.004 Peters-Lidard, C. D., Houser, P. R., Tian, Y., Kumar, S. V., Geiger, J., Olden, S., Lighty, L., Doty, B., Dirmeyer, P., Adams, J., Mitchell, K., Wood, E. F., & Sheffield, J. (2007). High-performance Earth system modeling with NASA/GSFC's Land Information System. Innovations in Systems and Software Engineering, 3(3), 157–165. https://doi.org/10.1007/s11334-007-0028-x Sohoulande, C. D., Martin, J., Szogi, A., & Stone, K. (2020). Climate-driven prediction of land water storage anomalies: An outlook for water resources monitoring across the conterminous United States. Journal of Hydrology, 588, 125053. Rodell, M., Houser, P. R., Jambor, U. E. A., Gottschalck, J., Mitchell, K., Meng, C. J., ... & Toll, D. (2004). The global land data assimilation system. Bulletin of the American Meteorological society , 85 (3), 381-394. Additional Declarations No competing interests reported. Supplementary Files 0SuppFinal0922.docx Cite Share Download PDF Status: Published Journal Publication published 24 Mar, 2026 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 02 Jan, 2026 Reviews received at journal 30 Nov, 2025 Reviewers agreed at journal 01 Nov, 2025 Reviews received at journal 31 Oct, 2025 Reviewers agreed at journal 20 Oct, 2025 Reviewers invited by journal 06 Oct, 2025 Editor assigned by journal 06 Oct, 2025 Editor invited by journal 06 Oct, 2025 Submission checks completed at journal 26 Sep, 2025 First submitted to journal 26 Sep, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7698407","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":530709906,"identity":"7fec0045-7a61-463d-898d-25a97f3fb7fb","order_by":0,"name":"Behshad Mohajer","email":"","orcid":"","institution":"Arizona State University","correspondingAuthor":false,"prefix":"","firstName":"Behshad","middleName":"","lastName":"Mohajer","suffix":""},{"id":530709907,"identity":"a9b8e165-f6a4-4c62-b222-aada09dec927","order_by":1,"name":"James S. Famiglietti","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABDUlEQVRIiWNgGAWjYBACxgYgUcHAwMPADmZKyEHE2QhoOQPSwgxkHkiQMCaoBQzOgAhmID6QwJDYQEgLc/sZA4YDNdtk5J2ZGx9//GGR3t8PFPlQdhi3w3pygFqO3eYxPMzYbAB0WO6MGzkGjDPO4dHSkGPA/IENqKWZsU0CpKXhBo8BM28bHi39b4C2/ANraf8B1JIuf/6MAfNffFpmAB12sO02jzwzYxsoxBIMDgDtZcSr5VnBgYN9t4GOYWyWOJMmYbjxRlrBwZ5z6Ti1GPYnb3xw4Ntte/n29ocfKmzq5OXOH9744EeZNW4tDRwGB0AMCAkFB7AphQF5BvYHEEYDPmWjYBSMglEwogEA0jlcsA/55E8AAAAASUVORK5CYII=","orcid":"","institution":"Arizona State University","correspondingAuthor":true,"prefix":"","firstName":"James","middleName":"S.","lastName":"Famiglietti","suffix":""},{"id":530709908,"identity":"e04514c4-14a2-4093-a9eb-36da86c26384","order_by":2,"name":"Hrishikesh A. Chandanpurkar","email":"","orcid":"","institution":"Arizona State University","correspondingAuthor":false,"prefix":"","firstName":"Hrishikesh","middleName":"A.","lastName":"Chandanpurkar","suffix":""},{"id":530709909,"identity":"99cb33ef-6fde-491f-b299-7eda00af1101","order_by":3,"name":"Fengwei Hung","email":"","orcid":"","institution":"Arizona State University","correspondingAuthor":false,"prefix":"","firstName":"Fengwei","middleName":"","lastName":"Hung","suffix":""}],"badges":[],"createdAt":"2025-09-24 01:53:18","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7698407/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7698407/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-026-44132-0","type":"published","date":"2026-03-24T16:10:08+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":93764278,"identity":"eacf3535-814e-4ec6-8520-a89883d13fe8","added_by":"auto","created_at":"2025-10-17 10:20:47","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2478087,"visible":true,"origin":"","legend":"","description":"","filename":"0ScientificReportsFinal0926.docx","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/babf82dc9eaf869dbe12d963.docx"},{"id":93765479,"identity":"d45620a8-844f-442d-a7ad-f094370ab1aa","added_by":"auto","created_at":"2025-10-17 10:28:47","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7151,"visible":true,"origin":"","legend":"","description":"","filename":"95e13f2056fc4a6e8864e12ae4c99fcd.json","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/738716764b61098bfbb8a6c0.json"},{"id":93764264,"identity":"89e58c94-b6c6-4a58-ae91-85accacc3181","added_by":"auto","created_at":"2025-10-17 10:20:47","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1390860,"visible":true,"origin":"","legend":"","description":"","filename":"0SuppFinal0922.docx","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/8ccc8bf10a2d820dcfec292c.docx"},{"id":93764262,"identity":"e743ed17-313e-4fef-9950-af57d9e4bdc1","added_by":"auto","created_at":"2025-10-17 10:20:47","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":157268,"visible":true,"origin":"","legend":"","description":"","filename":"95e13f2056fc4a6e8864e12ae4c99fcd1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/55a4bd21a67f119e955e8586.xml"},{"id":93765488,"identity":"ed9f6b91-3ec8-4174-836d-0dc8b31fcc58","added_by":"auto","created_at":"2025-10-17 10:28:47","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":992254,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/4975cd000ed27b0a1358853a.png"},{"id":93765482,"identity":"e2738552-8666-4326-9bdb-97c4ba0aa7b3","added_by":"auto","created_at":"2025-10-17 10:28:47","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":537339,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/b074ee9ba910cf45683ae3b8.png"},{"id":93765489,"identity":"5d64c233-d258-45af-926c-44d999dd576e","added_by":"auto","created_at":"2025-10-17 10:28:47","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":276359,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/e4db4d37a67aa779da078532.png"},{"id":93766560,"identity":"0fab50da-ab71-4da7-b566-95ec3da50afb","added_by":"auto","created_at":"2025-10-17 10:44:47","extension":"jpeg","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":251974,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/01346134b4f3a3b79fd3e286.jpeg"},{"id":93764268,"identity":"908c78a3-026c-47d8-9165-58c0545d91a6","added_by":"auto","created_at":"2025-10-17 10:20:47","extension":"jpeg","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":196974,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/f8563af0b48c515c251503d6.jpeg"},{"id":93765729,"identity":"efc33b1a-266d-40a2-a5ab-ec754209b565","added_by":"auto","created_at":"2025-10-17 10:36:47","extension":"jpeg","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":134973,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/281083279f5673e2709ea1a1.jpeg"},{"id":93764270,"identity":"f9d2e646-46da-429a-953d-9ccc72e99f04","added_by":"auto","created_at":"2025-10-17 10:20:47","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":151064,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/59c4f7e4f1cfd7fabe5ebbc8.png"},{"id":93765728,"identity":"e09a076c-8f14-4aa6-b226-1bb423237793","added_by":"auto","created_at":"2025-10-17 10:36:47","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":75542,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/eb4ebda22b54549c2bb26051.png"},{"id":93764273,"identity":"966d62ae-5f32-4736-85eb-5b18c4816ae5","added_by":"auto","created_at":"2025-10-17 10:20:47","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":45148,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/91d3108af9f342e6189a3157.png"},{"id":93764266,"identity":"363fa29f-6b43-46f7-8b83-8b95d448d46f","added_by":"auto","created_at":"2025-10-17 10:20:47","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":81646,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/892dc98b1239e20009a5f662.png"},{"id":93765730,"identity":"3a8d5024-e3d5-4489-9416-cb84c52058c3","added_by":"auto","created_at":"2025-10-17 10:36:47","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":63374,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/d1ef28a6b59535847fd6df76.png"},{"id":93764274,"identity":"710a2074-0739-4f8b-b40d-0a0c48b583e0","added_by":"auto","created_at":"2025-10-17 10:20:47","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":60855,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/f0b9934d1564588e6d5ea256.png"},{"id":93764275,"identity":"e22b0a98-acb4-45e8-9457-ed19e69b73ec","added_by":"auto","created_at":"2025-10-17 10:20:47","extension":"xml","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":155459,"visible":true,"origin":"","legend":"","description":"","filename":"95e13f2056fc4a6e8864e12ae4c99fcd1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/87fd76eb9b02145ae33aa35e.xml"},{"id":93765490,"identity":"eeea9ea7-8c46-4de5-9648-3a68de0a1cac","added_by":"auto","created_at":"2025-10-17 10:28:47","extension":"html","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":170136,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/2c4616bdb62e30ed8035cf91.html"},{"id":93765478,"identity":"2f45a815-db84-4421-81ef-659da6cb9c1a","added_by":"auto","created_at":"2025-10-17 10:28:46","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":406524,"visible":true,"origin":"","legend":"\u003cp\u003eSpatial distribution of standardized linear trends in GRACE/FO-derived Groundwater Storage Anomalies (GWSA) and WLDAS hydroclimatic variables across selected Arizona subbasins (2004–2021). For each variable, pixel-wise trends were computed using Ordinary Least Squares (OLS) regression of monthly values, spatially masked by subbasin boundaries, and standardized to enable direct comparison. A standard color scale highlights relative hotspots of increasing (blue) and decreasing (red) standardized trends across variables, with subbasin boundaries shown in black and statewide groundwater subbasins in gray. Trends estimated OLS on monthly standardized anomalies, 2004–2021; mask: study subbasins; grid: 0.25°\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/27c6e5d63ebf0b9663ebf4df.jpg"},{"id":93764254,"identity":"ea5d17cc-7ef8-4d3d-8c0f-b7fe3c81185d","added_by":"auto","created_at":"2025-10-17 10:20:46","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":190632,"visible":true,"origin":"","legend":"\u003cp\u003eStandardized correlations between GRACE/FO-derived groundwater storage anomalies (GWSA) and selected WLDAS hydrometeorological variables across central and southern Arizona groundwater subbasins for 2004 to 2021.\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/de430b85f3d9722673202502.jpg"},{"id":93765480,"identity":"90dd8152-06e2-4bfe-88b0-a8eb166c1291","added_by":"auto","created_at":"2025-10-17 10:28:47","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":119183,"visible":true,"origin":"","legend":"\u003cp\u003ePCA of the subbasin-wise correlation matrix) yields three interpretable modes explaining 94.3% of variance. PCA explains 94.3% of variance (PC1 80.9%, PC2 8.2%, PC3 5.1%)\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/b6f670ea308f802540e4b94b.jpg"},{"id":93764260,"identity":"2f590ba2-f5d1-46fa-84cb-a72784b5502d","added_by":"auto","created_at":"2025-10-17 10:20:47","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":153684,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eSpatial distribution of PCA-based hydroclimatic clusters for central and southern Arizona subbasins (2004–2021). Clusters were derived from k-means classification of subbasin scores from the first three principal components linking GRACE-derived groundwater storage anomalies (GWSA) with WLDAS hydrometeorological variables.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/74e3ea36b9db9740af25502f.jpg"},{"id":93765481,"identity":"7fb4f075-d6d3-40ea-ad39-94ab224cef6f","added_by":"auto","created_at":"2025-10-17 10:28:47","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":155503,"visible":true,"origin":"","legend":"\u003cp\u003eA map of the study area in central and southern Arizona showing subbasins selected for the present analysis. Subbasins included in the analysis are from the Arizona Department of Water Resources (ADWR) (https://gisdata2016-11-18t150447874z-azwater.opendata.arcgis.com/ ).\u0026nbsp;\u003c/p\u003e","description":"","filename":"5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/687ab84443e406766e20ab80.jpg"},{"id":93764261,"identity":"88faef30-c16b-4a3b-8cd0-1b51d5ce4e90","added_by":"auto","created_at":"2025-10-17 10:20:47","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":106364,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eA schematic overview of the workflow is used to diagnose hydroclimatic controls on groundwater storage. GRACE/FO GWSA (0.25°) and WLDAS variables (0.01°) are pre-processed by co-gridding and converting to standardized monthly anomalies (2004–2021). Pearson correlations between GWSA and each WLDAS variable are computed and averaged by groundwater subbasin to form a subbasin-by-variable matrix. Principal component analysis is then applied to this matrix, and the resulting PC scores are clustered with k-means to delineate hydroclimatic regimes across central and southern Arizona.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/75a48b2e3315e2b1a0386ee1.jpg"},{"id":105756079,"identity":"5055b30a-1571-413b-a59d-27e13b42515a","added_by":"auto","created_at":"2026-03-30 16:35:17","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1780623,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/a6a6c7ee-c065-423f-9673-3cd15615225f.pdf"},{"id":93764256,"identity":"a815c712-b1f2-438b-a191-cd942107d341","added_by":"auto","created_at":"2025-10-17 10:20:47","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":1390860,"visible":true,"origin":"","legend":"","description":"","filename":"0SuppFinal0922.docx","url":"https://assets-eu.researchsquare.com/files/rs-7698407/v1/98a10820283208ad5cc924f3.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Key natural influences on groundwater storage changes in central and southern Arizona","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eThe Colorado River Basin (CRB) encompasses nearly 250,000 square miles of the arid and semi-arid American Southwest, draining portions of seven western US states and northern Mexico [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. The CRB is hydrologically divided by the Lees Ferry, Arizona. The Upper Colorado River Basin (UCRB) delivers approximately 90% of the total flow into the CRB. At the same time, the Lower Colorado River Basin (LCRB) of Arizona, California, and Nevada, which is home to most of the urban and agricultural areas, contributes over 55% of the consumptive use [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. This major river system supports over 40\u0026nbsp;million people, irrigates over five million acres of farmland, and supports a diverse set of ecosystems and economies in the region [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. The UCRB encompasses a unique geological setting of fractured sedimentary formations with confined aquifers. In contrast, the LCRB geologic setting features deep alluvial aquifers comprised of inter-bedded sands and silts with varying connectivity [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. These distinct structures generate highly non-uniform rates of recharge, groundwater-surface water interactions, and aquifer reactions to stress, which are increasingly being tested by a historic megadrought [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. The most extreme drought in 1,200 years accelerates hydrological imbalances across the CRB [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Declining precipitation, snowpack, reservoir storage, and rising atmospheric demand have significantly reduced natural recharge in the basin [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. On the other hand, groundwater is increasingly being used to compensate for shortfalls in surface water, especially in central and southern Arizona [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. The potential reduction of Colorado River allocations under post-2026 guidelines only heightens the urgency for groundwater resilience, particularly in central and southern Arizona [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eCentral and Southern Arizona - particularly the subbasins in and around the Active Management Areas (AMAs) -have a unique place in the LCRB where regional hydrological pressures combine with diverse geological and water use patterns [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. This region is in the Basin and Range Province, an arid region that contains deep, unconsolidated alluvial aquifers in fault-bounded basins with large amounts of water stored (see section 5.1). Still, accessibility varies throughout the region [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Recharge in these aquifers relies on episodic infiltration in mountain fronts and along flowing river channels, resulting in heterogeneous responses to natural and anthropogenic drivers [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Groundwater was historically a primary water source in Central Arizona until the Central Arizona Project (CAP) was built in the 1980s, allowing for large-scale imports of Colorado River Water [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. Rapid urban growth, multi-year drought, and surface water cutbacks have recently renewed reliance on groundwater in the region [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. While regulation and Managed Aquifer Recharge (MAR) programs like the Arizona Water Banking Authority (AWBA) have improved aquifer depletion, groundwater declines continue\u0026mdash;particularly outside of regulated areas\u0026mdash;because of pumping overdraft, limited natural recharge, and variable spatially specific management [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. This highlights the importance of diagnosing assessments that incorporate surface climate variability, aquifer response variability, and human management signals to determine vulnerabilities and inform regional groundwater planning with a view to sustainability [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eSatellite gravimetry has dramatically increased the potential for large-scale groundwater monitoring [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. NASA's Gravity Recovery and Climate Experiment (GRACE) and its successor GRACE-Follow On (GRACE/FO) (to refer to the combined GRACE and GRACE-FO from here on) have allowed regular observation of changes in total Terrestrial Water Storage (TWS) since they launched in 2002, through the measurement of tiny shifts in Earth's gravity field [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. Groundwater Storage Anomalies (GWSA) are calculated by removing observations or model simulations of snow, soil moisture, and surface water from TWS, providing monthly measures of downward integrated groundwater change rates [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. The global coverage and consistency of GRACE/FO are particularly advantageous in regions where very few in situ monitoring sites exist [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. However, although GRACE/FO provides reliable measures of water storage changes, it cannot identify the surface processes by which that change occurs [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. Several recent studies indicate that combining GRACE/FO with land surface models improves the capacity to diagnose groundwater trends and climate drivers related to those trends [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e]. Few of these studies have facilitated actionable insights for practitioners in planning contexts. Researchers have increasingly turned to land surface models like NASA's Western Land Data Assimilation System (WLDAS) [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. This simulates physical processes across the land-atmosphere interface [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e]. WLDAS is built upon the Noah-MP Land Surface Model [\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e] and incorporates satellite and in-situ data to simulate surface fluxes such as evapotranspiration, runoff, and snowmelt, and storages such as soil moisture, with high spatial and temporal scales [\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e]. WLDAS offers a consistent and high-resolution view of surface conditions influencing groundwater recharge, especially in data-scarce and topographically diverse regions like Arizona.\u003c/p\u003e\u003cp\u003eWe present a diagnostic approach integrating satellite-based groundwater observations with land surface model data to assess subbasin-scale hydroclimatic controls on groundwater storage. We test our approach in central and southern Arizona region. By combining GRACE/FO-derived GWSA with WLDAS hydrometeorological variables, this study captures both the spatial variability and temporal dynamics of surface\u0026ndash;subsurface interactions across Central and Southern Arizona. Our approach centers on three core steps: (1) quantifying long-term trends in groundwater and surface conditions using standardized anomalies, (2) identifying key climate\u0026ndash;groundwater linkages through correlation analysis, and (3) extracting dominant hydroclimatic modes via Principal Component Analysis (PCA). We further classify subbasins into similar groups using clustering of PCA scores to enhance interpretability and support spatial management planning. This multi-method approach enables a robust assessment of how groundwater variability responds to climate and offers practical insight into regional groundwater vulnerability under future water stress scenarios.\u003c/p\u003e"},{"header":"2. Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e2.1. Spatial patterns of long-term trends\u003c/h2\u003e\u003cp\u003eThe spatial distributions of standardized linear trends in GRACE/FO-derived GWSA and WLDAS surface variables from 2004 to 2021 (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) show pronounced hydroclimatic heterogeneity in the study area. These trends reflect the combined effects of atmospheric forcing, surface water fluxes, and storage processes, indicating distinct climate\u0026ndash;water trajectories across subbasins.\u003c/p\u003e\u003cp\u003eThe southern subbasins show increased surface temperature, air temperature, total evapotranspiration, and vegetation transpiration, signaling a sustained increase in atmospheric water demand and land surface energy flux. This also corresponds with declines in deep soil moisture. The simultaneous increases in evapotranspiration and decreases in all the storage variables lead to a net groundwater loss trajectory - in line with trends observed in other arid and semi-arid basins experiencing climate-driven groundwater decline [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e].\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eIn comparison, the northern and central subbasins show a positive trend in both rainfall rate and deep soil moisture, with limited increases in surface and subsurface runoff. These hydrologic anomalies suggest increased infiltration and possible recharge. In these basins, the GRACE/FO GWSA were generally neutral to mildly positive, suggesting the combination of improved moisture inputs and the regional Managed Aquifer Recharge (MAR) programs in some areas, such as those operated by Central Arizona Groundwater Replenishment District (CAGRD) [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eRunoff patterns exhibit substantial spatial variability. For surface runoff, positive trends are confined to the northern and part of the central basins with increasing precipitation and humidity. The trends in subsurface runoff show a similar pattern but are less consistent; nevertheless, there are significant increases in parts of the north and isolated parts of the central basins.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e2.2. Correlation of GRACE/FO GWSA and surface fluxes\u003c/h2\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e shows how GRACE/FO-derived GWSA and selected WLDAS hydroclimatic variables correlate at the subbasin level. This provides a diagnostic view of the relationship between natural surface\u0026ndash;atmosphere processes and groundwater variability throughout the study area. The positive correlations between GRACE/FO GWSA and precipitation rate, deep soil moisture, and subsurface runoff are most pronounced in places in the northern and some central subbasins. In these areas, storage changes are closely related to hydrologic inputs and provide evidence for recharge-dominated processes, in which seasonal to interannual climate variability directly impacts subsurface water storage. This correlation is consistent with conditions in relatively undeveloped basins or those with managed recharge facilities that enhance infiltration and deep percolation [\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eOn the other hand, southern subbasins show high negative standardized correlations between GRACE/FO GWSA and evapotranspiration, vegetation transpiration, and surface runoff. These subbasins are an example of loss-driven regimes, with strong atmospheric water demand, large land surface energy flux, and limited infiltration capacity; the inverse relationship between storage and surface fluxes shows vulnerability to increased evaporative stress and potential for continued aridification [\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e]. The inverse relationship between storage and surface fluxes reflects vulnerability to evaporative stress and progressive aridification, which are unlikely to reverse without supplemental recharge or reduced extraction [\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e]. Interpreting these correlations requires careful consideration of differences in data composition. GRACE/FO GWSA anomalies represent total terrestrial water storage changes, integrating natural hydroclimatic variability and anthropogenic influences such as pumping, artificial recharge, and aquifer management. WLDAS variables, by contrast, are outputs from a physically based land surface model that represents only the natural water and energy balance, excluding human water use [\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. This distinction is crucial in AMAs\u0026mdash;including Phoenix, Pinal, and parts of Tucson\u0026mdash;where artificial recharge from CAP deliveries contributes substantially to storage. In such basins, positive GRACE/FO trends or weak correlations with natural variables may be partly driven by Managed Aquifer Recharge (MAR) or other human interventions rather than climate inputs alone (See supplemental information Figure S7).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eAdditionally, to test against serial correlation, we computed cross-correlations at lags \u0026minus;\u0026thinsp;3 to +\u0026thinsp;3 months for the 24 target subbasins and re-tested zero-lag significance using an AR(1) effective degrees-of-freedom adjustment. Median patterns indicate that evapotranspiration and vegetation transpiration tend to lead GWSA by about one to two months; rain rate and surface runoff also show the highest at positive lags (variables leading GWSA by roughly one to two months); deep soil moisture peaks near zero lag; and subsurface runoff remains positive across lags. Because GWSA is strongly persistent, the degree of freedom correction reduces nominal significance, yet a non-trivial subset of basins retains detectable zero-lag coupling after adjustment. These coherent lead\u0026ndash;lag structures support our variance-attribution result that natural modes account for about one-sixth of spatial trend differences, with evapotranspiration contributing the largest share. Complete distributions, heatmaps, autocorrelation histograms, and significance summaries are provided in the Supplementary Information text S3 and figures S6 to S8.\u003c/p\u003e\u003cp\u003eThese findings highlight the importance of integrating human water-use datasets in future analysis to partition the respective contributions of climate forcing and anthropogenic management. Without this integration, correlations between GRACE/FO GWSA and natural variables risk conflating climate\u0026ndash;groundwater coupling with infrastructure-driven storage changes, especially in regions where policy and management interventions strongly mediate the hydrologic signal. Because GRACE/FO integrates human and natural signals, whereas WLDAS represents only natural processes, correlation strength should not be interpreted as causation, particularly within AMAs, where MAR and pumping can decorrelate climate-storage coupling. Conversely, anthropogenic withdrawals can dampen or reverse the natural correlation signal in heavily pumped basins without significant artificial recharge. For example, in agricultural zones with high extraction rates and limited water imports, GRACE/FO GWSA may show declining trends even where WLDAS variables indicate favorable natural recharge conditions [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e].\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e2.3. Principal Component Analysis (PCA)\u003c/h2\u003e\u003cp\u003ePC1 loads with similar magnitude and sign across all six variables (see tables S1 and S2), representing a coherent hydro-climate coupling mode. PC2 separates atmospheric flux coupling from runoff/plant-use pathways, while PC3 emphasizes deep-soil storage (SM100\u0026ndash;200) opposed by plant use/quick runoff. These patterns are spatially consistent with correlation maps, delineating a recharge-responsive north/central tier and a loss-dominated southern tier, with AMA basins showing management-modulated signals.\u003c/p\u003e\u003cp\u003ePCA score maps (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) show coherent regional groupings across the mapped subbasins (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). PC1 is low in basins such as the East Salt River Valley (ESR), West Salt River Valley (WSR), and Harquahala (HAR). At the same time, it is high in Aguirre Valley (AGV) and Santa Cruz AMA (SCA), with Avra Valley (AVR) and Maricopa\u0026ndash;Stanfield (MST) positive to near neutral. PC2 separates atmospheric-flux versus runoff/plant-use pathways scores that are low/negative in HAR, Gila Bend (GIL), Agua Fria (AGF), and positive in SCA. PC3 emphasizes deep-soil storage versus plant-use/quick-runoff. Scores are near zero in ESR, modestly positive in MST, Eloy (ELO), Vekol Valley (VEK), and Gila Bend (GIL), and higher positive in Santa Rosa Valley (SRO).\u003c/p\u003e\u003cp\u003eWe also examined how long-term groundwater storage changes relate to identified PCA-derived hydroclimatic modes. Regression of basin trends from GRACE/FO-derived subbasins onto the first three PCs explained about 16% of the spatial variance, which indicates natural variability explained a small but measurable part of the observed differences between subbasins. Among the explained variance, total evapotranspiration had the most significant contribution (\u0026asymp;\u0026thinsp;28.6%), followed by rainfall (\u0026asymp;\u0026thinsp;23.4%) and subsurface runoff (\u0026asymp;\u0026thinsp;19.8%), while vegetation transpiration (\u0026asymp;\u0026thinsp;13.0%), surface runoff (\u0026asymp;\u0026thinsp;10.0%), and deep soil moisture (\u0026asymp;\u0026thinsp;5.2%) had smaller contributions. These values are consistent with Arizona's hydroclimatic setting. High evaporative demand is responsible for a significant share of water loss, precipitation, and mountain-front recharge as essential inputs, and deep infiltration signals are degraded and less pervasive at the basin scale.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eAcross the study area, recharge-related processes accounted for approximately 58% of the explained variance compared with loss processes, which accounted for 42%. Given the modest R\u0026sup2;, these shares should be viewed as directional rather than definitive. This reflects the broader theme that spatial variability in recharge potential explains why specific subbasins experience slower decline rates than others. For management purposes, this brings attention to the need to protect recharge corridors and/or maximize opportunities for recharge (infiltration) in basins where recharge is favored; however, it acknowledges that losses due to the atmosphere and water use will continue to constrain resilience in hotter and drier subbasins operatively.\u003c/p\u003e\u003c/div\u003e"},{"header":"3. Discussion","content":"\u003cp\u003eThis study comprehensively evaluates hydroclimatic controls on groundwater storage variability in Central and Southern Arizona, combining GRACE/GRACE-FO satellite gravimetry observations, WLDAS-modeled surface fluxes, and PCA-based dimensionality reduction. The results offer new insights into the spatial heterogeneity of groundwater\u0026ndash;climate interactions and reveal distinct hydroclimatic modes that help explain observed depletion patterns, resilience, and management vulnerability.\u003c/p\u003e\u003cp\u003ePrincipal Component Analysis results (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) indicate that subbasins do not respond uniformly to climate forcing. PC1 is a hydro-climate coherence mode reflecting the overall strength of common coupling between GWSA and precipitation, soil-moisture, runoff, and ET, not a recharge\u0026ndash;loss gradient. The recharge\u0026ndash;loss contrast emerges on PC2 and PC3: PC2\u0026rsquo;s sign distinguishes an atmospheric-flux pathway (higher/positive PC2: ET and P side) from a runoff/plant-use pathway (lower/negative PC2: Qsb, Qs, and Tveg). In contrast, higher PC3 indicates stronger deep-soil-storage control. Higher PC3 (and the PC2 sign) mark more recharge-responsive behavior\u0026mdash;often amplified where managed aquifer recharge is present. In contrast, low/negative PC2 and/or low PC3 indicate loss-dominated regimes in which atmospheric demand and plant water use/quick runoff decouple GWSA from recharge. Subbasins in this latter group also show negative GWSA trends and weak or negative correlations with recharge variables, reinforcing their vulnerability to hydrologic stress.\u003c/p\u003e\u003cp\u003eOne of the key challenges in interpreting GRACE/FO-based groundwater anomalies lies in disentangling natural hydroclimatic variability from anthropogenic influences such as pumping and artificial recharge. While the PCA is based on correlations between GRACE/FO (which includes both human and natural signals) and WLDAS (which models only natural processes), mismatches in certain AMAs\u0026mdash;such as Phoenix and Pinal\u0026mdash;highlight the limitations of relying on climate data alone. These patterns likely reflect intensive water management, where Central Arizona Project (CAP) imports, recharge basins, and regulatory frameworks mask the true extent of climatic stress. The divergence between GRACE/FO and WLDAS in such regions reinforces the need for dual attribution frameworks that separately analyze natural and anthropogenic controls. Without accounting for managed recharge, models may underestimate the buffering effect of infrastructure. Similarly, excluding pumping data prevents a complete diagnosis of overdraft severity. Future studies could incorporate modeled or observed groundwater extraction datasets or use inversion techniques to estimate human extraction as a residual from the GRACE/FP\u0026ndash;WLDAS signal mismatch.\u003c/p\u003e\u003cp\u003eClustering of the subbasin PC scores partitions the region into four hydroclimatic regimes (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Cluster 4 (dark green) dominates the southern tier\u0026mdash;covering most of the agricultural valleys\u0026mdash;and represents loss-dominated, climate-sensitive behavior (low/negative PC3 and low PC2). Cluster 3 (blue) forms a south-central belt, indicating mixed controls where atmospheric demand and limited recharge both matter and where management signals are likely. Cluster 2 (green) is concentrated along the north-central, higher-relief corridor, reflecting more recharge/storage-responsive behavior (higher PC2/PC3). Cluster 1 (light) appears in a few central/western rim basins and represents weaker or transitional coupling. This segmentation is consistent with the well-trend and GRACE patterns\u0026mdash;Cluster 4 aligns with widespread declines, Clusters 2\u0026ndash;3 include more stable or improving conditions, and provide a management-ready way to target pumping curbs, recharge expansion, and monitoring effort.\u003c/p\u003e\u003cp\u003eThese findings can inform groundwater governance under Arizona's post-2026 water management landscape. With Colorado River shortages likely to persist, reliance on groundwater will intensify, particularly in agriculture-heavy areas like Pinal AMA [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. The PCA and clustering results highlight which subbasins are least capable of absorbing additional stress without crossing sustainability thresholds. A key policy takeaway is the need for spatially differentiated management. Instead of uniform restrictions or recharge targets across AMAs, management strategies are more effective if they reflect each subbasin's hydroclimatic regime and recharge potential. For instance, basins in Cluster 4 may require stricter pumping limits or subsidized artificial recharge, while Cluster 1 basins may benefit more from preserving infiltration zones or incentivizing natural recharge.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eSubbasin classification based on hydroclimatic drivers reveals clear management priorities across Central and Southern Arizona. Loss-dominated basins are highly climate-sensitive, with low natural recharge and high evaporative demand, requiring long-term conservation and Managed Aquifer Recharge (MAR) support. Mixed-stress basins face moderate recharge potential and intense pumping pressure, highlighting the need for tighter integration of surface water, climate, and groundwater management. Recharge-responsive areas show favorable conditions for infiltration and offer strategic opportunities to expand recharge programs. In contrast, compound-risk basins like Gila Bend are more vulnerable, climate-stressed, and unregulated. Gila Bend exemplifies the risks of unmanaged extraction outside AMA boundaries. These results underscore the need for an adaptive, risk-based groundwater policy addressing natural vulnerability and institutional gaps.\u003c/p\u003e\u003cp\u003eThe discussion of PCA and clustering results is compared to well-level groundwater trends observed by the Arizona Department of Water Resources (ADWR) (See text S2 and figures S2, S3, and S4). Areas identified as climatically sensitive and loss-dominated correspond with depleted subbasins, indicating long-term groundwater decline. In contrast, basins with high PC1 scores and high recharge correlations show fewer over-drafted wells, supporting their classification as more resilient. In-situ groundwater monitoring using data collected from ADWR index wells (located in unconfined aquifers) shows substantial differences in spatial and management-based depth-to-water trends across central and southern Arizona from 2000 to 2023. Subbasin locations in the south and western areas of the study demonstrate extensive and steep declines (red shading in Figure S3; red wells in Figure S4), reflecting groundwater intensive use and low natural recharge.\u003c/p\u003e\u003cp\u003eConversely, groundwater conditions in some northern and eastern subbasins have shown relatively stable or rising circumstances (blue shading in Figure S4; blue wells in Figure S5), frequently correlated with managed recharge activities and lower groundwater withdrawal pressure. Time-series analysis (Figure S5) confirmed notable, statistically significant managed groundwater rising levels AMAs and declining levels in non-AMAs, confirming groundwater governance plays a role in securing declines in water levels. Nonetheless, declining water levels in parts of the AMAs suggest that application of regulatory protections is insufficient unless protection measures include interventions to address climate-driven pressures and groundwater extraction levels.\u003c/p\u003e\u003cp\u003eThe comparison between this study\u0026rsquo;s results and the latest AMA\u0026rsquo;s Management Plans (4MPs) (available online at: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.azwater.gov/fourth-management-plan\u003c/span\u003e\u003cspan address=\"https://www.azwater.gov/fourth-management-plan\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) shows that management plans may only align with at least partially the hydroclimatic trends outlined by GRACE/FO and WLDAS. There are identified recharge-resilient basins in the northern and central best-practice basins, and those have associated recharge and Assured Water Supply programs; however, these two are static and do not incorporate information from hydro-climatic variables. Transition basins, such as in the Santa Cruz and Pinal AMAs, use conservation programs to address possible issues, but use a fixed evapotranspiration value that does not address the changing climate stress. The intermediate basins remain stable but have weak evolving protection of recharge corridors or management on recovery siting.\u003c/p\u003e\u003cp\u003eThe Gila Bend subbasin is a significant case illustrating compounded vulnerability from natural and anthropogenic pressures. Gila Bend shows low PCA scores and high correlations with evaporative demand, indicating a natural hydroclimatic deficit. It also lies outside the boundaries of any AMA, leaving it largely exempt from pumping regulations. Long-term monitoring well data confirms steep groundwater declines, suggesting that this subbasin is exposed to climatic stress and unregulated extraction without large-scale artificial projects or management plans. This combination of natural and anthropogenic factors places the Gila Bend subbasin among the highest-risk basins in the study region. In short, while the 4MPs recognize recharge and conservation, they miss climate variability, ET change, and basin-specific needs, leaving room for more adaptive, science-based sustainability plans. This study offers a practical framework for prioritizing subbasins under the post-2026 water management scenario. Loss-dominated clusters may require stricter pumping regulation, expanded recharge infrastructure, and enhanced monitoring, whereas recharge-responsive clusters may benefit from preserving infiltration zones and sustaining artificial recharge operations.\u003c/p\u003e\u003cp\u003eAdditionally, the diagnostic approach presented here directly relates to physical and data-driven modeling approaches. In the case of process-based physical models, understanding the balance between natural hydroclimatic drivers and anthropogenic influences is essential to represent recharge, evapotranspiration, and storage change mechanisms accurately. In the Machine Learning (ML) and Artificial Intelligence (AI) models, the PCA structure provides an interpretable feature space that captures dominant patterns of groundwater\u0026ndash;climate coupling, reducing the risk of overfitting to noise. Without explicit knowledge of the natural and human processes shaping groundwater variability, physical models risk misrepresentation, and without identifying the leading principal components, ML models risk misattribution of drivers\u003c/p\u003e\u003cp\u003eSeveral sources of uncertainty must be acknowledged within the scope of our analysis. First, the GRACE/FO satellite data, while invaluable for detecting regional groundwater changes, are spatially coarse and reflect total Terrestrial Water Storage (TWS), which includes anthropogenic influences such as pumping and managed aquifer recharge. This aggregation makes it challenging to isolate purely natural signals without additional attribution methods. Second, the WLDAS dataset, while physically consistent and spatially detailed, models only natural processes and omits human water use, infrastructure, and policy interventions\u0026mdash;factors that are increasingly dominant in groundwater dynamics across central and southern Arizona. These model design and scope differences can introduce signal mismatches in correlation and PCA analyses, particularly in AMAs where human intervention is substantial. Moreover, the land surface model's performance depends on the choice of meteorological forcing and calibration datasets, which can further influence simulated surface fluxes like evapotranspiration or runoff. To partially address these uncertainties, this study employed GRACE/FO products that assimilate observations and WLDAS outputs validated against streamflow and evapotranspiration.\u003c/p\u003e\u003cp\u003eFuture work can reduce uncertainty by integrating direct pumping data, MAR volumes, and socio-institutional variables. Such additions would enable a complete dual-attribution framework that distinguishes climate-driven variability from anthropogenic impacts, thereby enhancing the utility of satellite-model integration for groundwater governance. Extending this analysis with dynamical modeling\u0026mdash;such as groundwater flow or water balance simulations\u0026mdash;would strengthen causal attribution and scenario planning. Applying PCA in moving windows (e.g., 5-year intervals) may reveal shifts in dominant controls as climate stress and policy responses unfold.\u003c/p\u003e"},{"header":"4. Conclusion","content":"\u003cp\u003eThis study presents an integrated assessment of the hydroclimatic controls causing groundwater storage variability across central and southern Arizona through the integration of GRACE/FO satellite gravimetry observations, hydrometeorological products from the Western Land Data Assimilation System (WLDAS), and multivariate statistical methods (e.g., Principal Component Analysis and k-means clustering).\u003c/p\u003e\u003cp\u003eThe loss-dominated subbasins feature groundwater trends that are consistently negative over time, and a strong relationship with evaporative fluxes and low PCA scores. These basins typically have little natural recharge potential, high atmospheric demand for water, and high sensitivity to warm and dry conditions made worse by climate effects, making them highly prone to groundwater depletion via climate stress. Meanwhile, we found that the recharge-responsive subbasins are highly associated with groundwater storage and natural recharge-related variables (i.e., rainfall, deep soil moisture, runoff), had higher PCA scores, and spatial clustering indicating higher resilience capacities. Extended artificial recharge schemes could enhance beneficial hydroclimatic preconditions in many of these basins. The regime map is diagnostic, not causal. It highlights where climate coupling is strong vs weak; targeted management data (pumping, MAR volumes) are needed to complete attribution.\u003c/p\u003e\u003cp\u003eThe diagnostic approach developed in this study is transferable to other regions with arid and semi-arid climate stress and changing surface water availability. Future work may incorporate human water-use metrics, artificial recharge volumes, and dynamic groundwater modelling to facilitate causal attribution efforts. PCA moving-window analysis could also discern shifts in subbasin sensitivity over timescales as conditions change and management actions occur.\u003c/p\u003e\u003cp\u003eAs Arizona and the Lower Colorado River Basin region manage more extended drought periods and changing surface water allocations, future water managers will need tools considering natural hydroclimatic variability and anthropogenic drivers to develop differentially focused, climate-informed, sustainability-based groundwater policies.\u003c/p\u003e"},{"header":"5. Methods","content":"\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e5.1. Study Area\u003c/h2\u003e\u003cp\u003eCentral and Southern Arizona are critical to evaluating groundwater vulnerability and sustainability in the Lower Colorado River Basin (LCRB) [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. The region contains the Phoenix, Pinal, and Tucson Active Management Areas (AMAs)\u0026mdash;bounded jurisdictions established to prevent groundwater overdraft and by the Arizona 1980 Groundwater Management Act [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. The AMAs are located within the ecologically arid to semi-arid Basin \u0026amp; Range Province. They are characterized by large, fault-contained, alluvial basins with accumulations of unconsolidated material\u0026mdash;with high potential for large-capacity aquifer systems, more prone to significant aquifer drawdown, subsidence, and degradation in response to an extended duration of natural and anthropogenic stressors [\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e, \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e, \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eCentral Arizona is a microcosm of layered water stress, where urban demand on a finite supply converges with irrigated agriculture, surface water availability diminished by human-caused climate change, and institutional constraints on supply and pumping [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e]. The surface water supply comes primarily from the Central Arizona Project (CAP) infrastructure, which is a 336-mile canal that delivers the State of Arizona's share of Colorado River water, and, to a lesser extent, local runoff from the Salt and Verde Rivers- not to mention the local micro-climate influences on the available surface water supply. However, CAP allocations are becoming more uncertain due to the prolonged megadrought, basin-wide over-allocation, and declining snowpack and runoff in the UCRB [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e]. Agricultural users are already increasingly dependent on groundwater for supply gaps, especially in places where CAP reductions have forced crop fallowing and re-drilling wells ever deeper [\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eRecharge in these basins is variable, typically low, precipitation-driven, streamflow-driven, and mountain-front driven [\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e]. Natural recharge rates are mostly low and spatially variable. However, managed recharge through basins and direct injections coordinated by the Arizona Water Banking Authority (AWBA) has become an essential means of groundwater recovery in AMAs where water may be available from CAP [\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e, \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e]. Not all subbasins benefit in equal degrees from such institutionally structured interventions; for example, regions of central Arizona have experienced over 100 meters of groundwater level decline and six meters of subsidence [\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e]. Therefore, focusing on Central and Southern Arizona presents an interesting testbed for scalable groundwater governance mechanisms, especially over the uncertain surface water demand and increasing frequency and intensity of climate extremes.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e5.2. Analytical workflow\u003c/h2\u003e\u003cp\u003eThis study employs a multi-step analytical workflow to quantify the influence of main hydroclimatic drivers on groundwater storage variability across groundwater subbasins in central and southern Arizona. The analysis consists of three significant steps: (1) anomaly calculation, standardization, and linear trend calculations, (2) correlation analysis, and (3) dimensionality reduction and clustering to identify the dominant hydroclimatic mode (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). For full methodological details, including equations and notation, see Supplementary Text S1\u0026ndash;S3.\u003c/p\u003e\u003cp\u003eAll the time series\u0026mdash;including GRACE/FO-derived groundwater storage anomalies (GWSA) and land surface variables from WLDAS\u0026mdash;were converted to standardized anomalies in the first stage. We formed standardized anomalies by subtracting the monthly climatology and dividing by the interannual standard deviation. Linear trends were then estimated on these standardized anomalies. Linear trend magnitudes were estimated via ordinary least squares regression at each 0.25\u0026deg; spatial grid cell [\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e], and slopes were subsequently standardized across the domain to allow spatial comparison of long-term changes in surface and subsurface water conditions. Next, the Pearson correlation coefficients were computed between standardized GRACE/FO anomalies and each WLDAS variable at the grid level [\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e]. These correlations were aggregated to the subbasin scale using zonal means, producing a subbasin-by-variable matrix that summarizes the strength and direction of hydroclimatic linkages across the study area. PCA was applied to the correlation matrix to extract the dominant modes of spatial variability [\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e]. The leading components, PC1 and PC2, identify the primary surface processes (e.g., evapotranspiration, runoff, soil moisture) influencing subbasin-scale groundwater variability.\u003c/p\u003e\u003cp\u003eSubbasins were embedded in a two-dimensional feature space defined by their PC1 and PC2 scores to identify spatial patterns in hydroclimatic control. K-means clustering [\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e] was then applied to classify subbasins into four groups with similar hydroclimatic signatures. The scikit-learn implementation of the K-Means algorithm was used with a fixed random seed to ensure reproducibility [\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e]. For further comparison, long-term trends in groundwater depth were estimated using in-situ ADWR Index Wells data across selected subbasins (see Supplementary Text S3 and Figures S4 and S5).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e5.3. Data\u003c/h2\u003e\u003cp\u003eThis study integrates satellite observations and model-derived datasets to examine the hydroclimatic controls on groundwater storage variability across central and southern Arizona. Specifically, we utilize (1) groundwater storage anomalies from the GRACE/FO missions, which are extracted from observations of Total Water Storage (TWS) [\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e], (2) land surface hydrometeorological variables from the Western Land Data Assimilation System (WLDAS) [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. These datasets collectively enable a multiscale assessment of groundwater and climate interactions from 2004 to 2021. Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e and Figure S2 describe the datasets used in this study.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eSummary of datasets\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\"\u0026times;\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eData\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSource\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eVariables\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSpatial Resolution\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eTemporal Span\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGRACE/FO Groundwater Storage Anomalies\u003c/p\u003e\u003cp\u003e(bias-corrected, enhanced resolution)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u0026bull; Jet Propulsion Laboratory (JPL)\u003c/p\u003e\u003cp\u003e\u0026bull; Chandanpurkar et al., 2025\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u0026bull; Total Water Storage Anomaly (TWSA)\u003c/p\u003e\u003cp\u003e\u0026bull; Groundwater Storage Anomaly (GWSA)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.25\u0026deg; \u0026times; 0.25\u0026deg;\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2004\u0026ndash;2021\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eWLDAS Land Surface Variables\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eWestern Land Data Assimilation System (WLDAS)\u003c/p\u003e\u003cp\u003e(Erlingis et al., 2021)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u0026bull; Surface Temperature (Avgsurft_Tavg),\u003c/p\u003e\u003cp\u003e\u0026bull; Air Temperature (Tair_F_Tavg),\u003c/p\u003e\u003cp\u003e\u0026bull; Specific Humidity (Qair_F_Tavg),\u003c/p\u003e\u003cp\u003e\u0026bull; Rain Precipitation Rate (Rainf_Tavg),\u003c/p\u003e\u003cp\u003e\u0026bull; Surface Runoff (Qs_Tavg),\u003c/p\u003e\u003cp\u003e\u0026bull; Subsurface Runoff (Qsb_Tavg),\u003c/p\u003e\u003cp\u003e\u0026bull; Total Evapotranspiration (Evap_Tavg),\u003c/p\u003e\u003cp\u003e\u0026bull; Vegetation Transpiration (Tveg_Tavg),\u003c/p\u003e\u003cp\u003e\u0026bull; Soil Moisture at 100\u0026ndash;200 Cm Depth (Soilmoi100_200cm_Tavg),\u003c/p\u003e\u003cp\u003e\u0026bull; Groundwater Storage (GWS_Tavg)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026times;\" colname=\"c4\"\u003e\u003cp\u003e0.01\u0026deg; \u0026times; 0.01\u0026deg;\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2004\u0026ndash;2021\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eClassification of central and southern Arizona subbasins.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"3\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCluster\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eHydroclimatic interpretation\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eManagement implications\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCluster 1 \u0026ndash; Recharge-responsive\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eIntermediate recharge signals, relatively neutral stress. PCA shows moderate positive coupling with soil moisture and precipitation, weak ET influence.\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eTransitional case: could respond positively to Managed Aquifer Recharge (MAR) expansion; needs monitoring to ensure pumping doesn\u0026rsquo;t tip balance toward depletion.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCluster 2 \u0026ndash; Recharge-dominated\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eHigh positive correlations of GWSA with precipitation, deep soil moisture, and subsurface runoff. Weak loss signals. PCA scores show high recharge coupling.\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNatural recharge potential is relatively high. Protect infiltration corridors (mountain fronts, ephemeral washes); MAR complements high natural signals.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCluster 3 \u0026ndash; Transition basins\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMixed recharge\u0026ndash;loss signatures; moderate soil-moisture/runoff coupling but elevated ET/transpiration stress. PCA scores intermediate. Many coincide with AMA urban/ag centers.\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eOutcomes hinge on management. climate stress. Long-term vulnerability if pumping \u0026gt;(natural\u0026thinsp;+\u0026thinsp;managed recharge). Proactive management plans are necessary.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCluster 4 \u0026ndash; Loss-dominated, high stress\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eWeak recharge signals; high negative correlations of GWSA with evapotranspiration and vegetation demand. PCA shows low/negative PC2\u0026ndash;PC3.\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eHighest compound risk: climate stress\u0026thinsp;+\u0026thinsp;weak regulation. Persistent Depth To Water (DTW) declines confirm loss-dominated status. Require stricter overdraft controls. MAR feasibility may be limited due to poor natural recharge capacity.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eWLDAS provides high-resolution, observation-driven estimates of key land surface hydrologic variables such as precipitation, evapotranspiration, soil moisture, and runoff. Developed to support drought monitoring and water resources research in the western United States. WLDAS integrates multiple remote sensing and in situ data sources with advanced land surface models to produce gridded outputs at 0.01\u0026deg; spatial resolution. While WLDAS offers valuable insights into land-atmosphere processes and captures hydroclimatic variability at fine spatial scales, uncertainties remain due to model structural limitations, errors in meteorological forcing inputs, and challenges in representing subsurface hydrology and anthropogenic water use. Despite these limitations, WLDAS datasets have been widely used for regional water balance studies and serve as a critical complement to satellite-based storage observations in this study [\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e, \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e, \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eThe GRACE/FO TWS data are based on the Landerer Mascon [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e] Release 6 Version 3 solution, which offers improved spatial resolution (0.25\u0026deg; \u0026times; 0.25\u0026deg;) through post-processing that integrates Global Land Data Assimilation System (GLDAS) models [\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e] and corrects for biases at the mascon scale [\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e]. This enhancement enables meaningful sub-mascon regional analysis while retaining consistency with the original GRACE observations. The processed dataset accounts for significant sources of uncertainty by incorporating JPL-provided formal errors and climatology-removed anomaly estimation.\u003c/p\u003e\u003cp\u003eTo estimate groundwater storage changes from GRACE/FO TWS, non-groundwater signals (e.g., snow water equivalent, surface and soil water storage) must be calculated and removed [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. This study utilizes global enhanced-resolution, bias-corrected groundwater storage anomaly data as processed and described by [\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e]. In our study, these data support the investigation of groundwater storage variability across central and southern Arizona, where hydrologic and climatic conditions require high-resolution satellite-informed datasets to assess groundwater\u0026ndash;climate interactions over the 2004\u0026ndash;2021 period.\u003c/p\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003ch2\u003eCompeting interests\u003c/h2\u003e\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003ch2\u003eAdditional information\u003c/h2\u003e\u003cp\u003eSupplementary Information\u003c/p\u003e\u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e\u003cp\u003eThis research was supported by funding from the Arizona State University School of Sustainability and College of Global Futures.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eJ.S.F. secured funding, provided resources, and supervised the project. B.M. and J.S.F. designed and conceptualized the study. B.M. performed the formal analysis, developed the software, carried out validation, and prepared the visualizations. B.M. and J.S.F. contributed to the investigation and methodology. J.S.F. led project administration. B.M. and J.S.F. wrote the original draft. All authors contributed to reviewing and editing the manuscript.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eMonthly hydrometeorological data were obtained from the NASA Western Land Data Assimilation System (WLDAS), specifically the Noah-MP Land Surface Model L4 daily product at 0.01\u0026deg; resolution ( [https://hydro1.gesdisc.eosdis.nasa.gov/data/WLDAS/WLDAS_NOAHMP001_DA1.D1.0/](https:/hydro1.gesdisc.eosdis.nasa.gov/data/WLDAS/WLDAS_NOAHMP001_DA1.D1.0) ) The data were accessed on November 4, 2024, via NASA's Earthdata system using authenticated download links. A bash automated script was used to retrieve and organize the NetCDF files locally for analysis. GRACE/GRACE-FO Mascon data are available at [https://podaac.jpl.nasa.gov/dataset/TELLUS_GRAC-GRFO_MASCON_CRI_GRID_RL06.3_V4](https:/podaac.jpl.nasa.gov/dataset/TELLUS_GRAC-GRFO_MASCON_CRI_GRID_RL06.3_V4) .\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eUS Bureau of Reclamation. (2012). Colorado River Basin Water Supply and Demand Study. https://www.usbr.gov/lc/region/programs/crbstudy/finalreport/Study%20Report/CRBS_Study_Report_FINAL.pdf\u003c/li\u003e\n\u003cli\u003eRichter, B. D., Lamsal, G., Marston, L., Dhakal, S., Sangha, L. S., Rushforth, R. R., Wei, D., Ruddell, B. L., Davis, K. F., Hernandez-Cruz, A., Sandoval-Solis, S., \u0026amp; Schmidt, J. C. (2024). New water accounting reveals why the Colorado River no longer reaches the sea. Communications Earth and Environment, 5(1). https://doi.org/10.1038/s43247-024-01291-0\u003c/li\u003e\n\u003cli\u003eFleck, J., \u0026amp; Udall, B. (2021). Managing Colorado River risk. Science, 372(6545), 885. https://doi.org/10.1126/SCIENCE.ABJ5498 \u003c/li\u003e\n\u003cli\u003eHunt, C., \u0026amp; Rabbitt, M. (1969). Geologic history of the Colorado River. In the US Geological Survey. https://pubs.usgs.gov/pp/0669/report.pdf#page=75 \u003c/li\u003e\n\u003cli\u003eGuay Jacobs, K. L., \u0026amp; Holway, J. M. (2004). Managing for sustainability in an arid climate: Lessons learned from 20 years of groundwater management in Arizona, USA. Hydrogeology Journal, 12(1), 52\u0026ndash;65. https://doi.org/10.1007/s10040-003-0308-y \u003c/li\u003e\n\u003cli\u003eHoward, K., \u0026amp; Bohannon, R. (2001). Lower Colorado River: Framework, neogene deposits, incision, and evolution. https://pubs.usgs.gov/publication/70198241\u003c/li\u003e\n\u003cli\u003eCrossey, L. C., Karlstrom, K. E., Dorsey, R., Pearce, J., Wan, E., Beard, L. S., Asmerom, Y., Polyak, V., Crow, R. S., Cohen, A., Bright, J., \u0026amp; Pecha, M. E. (2015). Importance of groundwater in propagating downward integration of the 6\u0026ndash;5 Ma Colorado River system: Geochemistry of springs, travertines, and lacustrine carbonates. Geosphere , 11, 660\u0026ndash;682. https://doi.org/10.1130/GES01073.1\u003c/li\u003e\n\u003cli\u003eWoodhouse, C. A., Pederson, G. T., Morino, K., McAfee, S. A., \u0026amp; McCabe, G. J. (2016). Increasing influence of air temperature on upper Colorado River streamflow. Geophysical Research Letters, 43(5), 2174\u0026ndash;2181. https://doi.org/10.1002/2015GL067613\u003c/li\u003e\n\u003cli\u003eSolander, K. C., Bennett, K. E., \u0026amp; Middleton, R. S. (2017). Shifts in historical streamflow extremes in the Colorado River Basin. Journal of Hydrology: Regional Studies, 12, 363\u0026ndash;377. https://doi.org/10.1016/j.ejrh.2017.05.004 \u003c/li\u003e\n\u003cli\u003eXiao, M., Udall, B., \u0026amp; Lettenmaier, D. P. (2018). On the Causes of Declining Colorado River Streamflows. Water Resources Research, 54(9), 6739\u0026ndash;6756. https://doi.org/10.1029/2018WR023153 \u003c/li\u003e\n\u003cli\u003eTran, H., Zhang, J., Cohard, J. M., Condon, L. E., \u0026amp; Maxwell, R. M. (2020). Simulating Groundwater-Streamflow Connections in the Upper Colorado River Basin. Groundwater, 58(3), 392\u0026ndash;405. https://doi.org/10.1111/gwat.13000 \u003c/li\u003e\n\u003cli\u003eBennett, K. E., Talsma, C., \u0026amp; Boero, R. (2021). Concurrent changes in extreme hydroclimate events in the Colorado River Basin. Water (Switzerland), 13(7). https://doi.org/10.3390/w13070978 \u003c/li\u003e\n\u003cli\u003eSchmidt, J. C., Yackulic, C. B., \u0026amp; Kuhn, E. (2023). The Colorado River water crisis: Its origin and the future. Wiley Interdisciplinary Reviews: Water, 10(6), e1672. https://doi.org/10.1002/WAT2.1672 \u003c/li\u003e\n\u003cli\u003eHoerling, M. P., Eischeid, J. K., Diaz, H. F., Rajagopolan, B., \u0026amp; Kuhn, E. (2024). Critical Effects of Precipitation on Future Colorado River Flow. Journal of Climate. https://doi.org/10.1175/JCLI-D-23-0617.1 \u003c/li\u003e\n\u003cli\u003eMartin, J. T., Pederson, G. T., Woodhouse, C. A., Cook, E. R., McCabe, G. J., Anchukaitis, K. J., ... \u0026amp; King, J. (2020). Increased drought severity tracks warming in the United States\u0026rsquo; largest river basin. Proceedings of the National Academy of Sciences, 117(21), 11328-11336. \u003c/li\u003e\n\u003cli\u003eGangopadhyay, S., Woodhouse, C. A., McCabe, G. J., Routson, C. C., \u0026amp; Meko, D. M. (2022). Tree Rings Reveal Unmatched 2nd Century Drought in the Colorado River Basin. Geophysical Research Letters, 49(11). https://doi.org/10.1029/2022GL098781 \u003c/li\u003e\n\u003cli\u003eMcCabe, G. J., Anchukaitis, K. J., Wise, E. K., Erger, P. J., Dolan, L., McGuire, M., Gangopadhyay, S., Chase, K. J., Littell, J. S., Gray, S. T., George, S. S., Friedman, J. M., Sauchyn, D. J., St-Jacques, J. M., \u0026amp; King, J. (2020). Increased drought severity tracks warming in the United States\u0026apos; largest river basin. Proceedings of the National Academy of Sciences of the United States of America, 117(21). https://doi.org/10.1073/pnas.1916208117\u003c/li\u003e\n\u003cli\u003ePokharel, B., Jagannathan, K. A., Wang, S. ‐Y. S., Jones, A., LaPlante, M. D., Buddhavarapu, S., Borhara, K., Ulrich, P., Leung, L. R., Eklund, J., Hasenyager, C., Serago, J., Prairie, J. R., Kaatz, L., Winchell, T., \u0026amp; Kugel, F. (2024). Can we rely on drought‐ending \u0026quot;miracles\u0026quot; in the Colorado River Basin? JAWRA Journal of the American Water Resources Association. https://doi.org/10.1111/1752-1688.13204\u003c/li\u003e\n\u003cli\u003eTillman, F. D., Gangopadhyay, S., \u0026amp; Pruitt, T. (2016). Changes in groundwater recharge under projected climate in the upper Colorado River basin. Geophysical Research Letters, 43(13), 6968\u0026ndash;6974. https://doi.org/10.1002/2016GL069714 \u003c/li\u003e\n\u003cli\u003eRyu, J., Wang, S.-Y., Jeong, J.-H., Kim, H., \u0026amp; Yoon, J.-H. (2025). Time of emergence (TOE) of potential aridification in the western United States. Journal of Hydrology, 626, 133029. https://doi.org/10.1016/j.jhydrol.2025.133029 \u003c/li\u003e\n\u003cli\u003eAbdelmohsen, K., Famiglietti, J. S., Ao, Y. Z., Mohajer, B., \u0026amp; Chandanpurkar, H. A. (2025). Declining Freshwater Availability in the Colorado River Basin Threatens Sustainability of Its Critical Groundwater Supplies. Geophysical Research Letters, 52(10). https://doi.org/10.1029/2025GL115593\u003c/li\u003e\n\u003cli\u003eScanlon, B. R., Pool, D. R., Rateb, A., Conway, B., Sorensen, K., Udall, B., \u0026amp; Reedy, R. C. (2025). Multidecadal drought impacts on the Lower Colorado Basin with implications for future management. Communications Earth and Environment, 6(1). https://doi.org/10.1038/s43247-025-02149-9 \u003c/li\u003e\n\u003cli\u003ePerry, D., Swanson, R. K., \u0026amp; Springer, A. E. (2024). Policy deficiencies and contingency plans: Groundwater management implications for baseflow contributions to the Colorado River. Frontiers in Environmental Science, 12, 1444015. https://doi.org/10.3389/fenvs.2024.1444015\u003c/li\u003e\n\u003cli\u003eTillman, F. D., \u0026amp; Leake, S. A. (2010). Trends in groundwater levels in wells in the active management areas of Arizona, USA. Hydrogeology Journal, 18(7), 1515\u0026ndash;1524. https://doi.org/10.1007/s10040-010-0603-3 \u003c/li\u003e\n\u003cli\u003eJacobs, K. L., \u0026amp; Holway, J. M. (2004). Managing for sustainability in an arid climate: lessons learned from 20 years of groundwater management in Arizona, USA. Hydrogeology Journal, 12(1), 52-65.\u003c/li\u003e\n\u003cli\u003eGarner, B. D., Pool, D. R., Tillman, F. D., \u0026amp; Forbes, B. T. (2013). Human effects on the hydrologic system of the Verde Valley, central Arizona, 1910\u0026ndash;2005 and 2005\u0026ndash;2110, using a regional groundwater flow model (Scientific Investigations Report 2013\u0026ndash;5029). US Geological Survey. https://doi.org/10.3133/sir20135029 \u003c/li\u003e\n\u003cli\u003eTillman, F. D., \u0026amp; Flynn, M. E. (2024). Arizona Groundwater Explorer: Interactive maps for evaluating the historical and current groundwater conditions in wells in Arizona, USA. Hydrogeology Journal, 32, 645\u0026ndash;661. https://doi.org/10.1007/s10040-023-02748-w \u003c/li\u003e\n\u003cli\u003eHanemann, W. M. (2002). The Central Arizona Project (Working Paper No. 937). Department of Agricultural and Resource Economics and Policy, Division of Agricultural and Natural Resources, University of California at Berkeley. Giannini Foundation of Agricultural Economics. https://escholarship.org/uc/item/9pd083d5 \u003c/li\u003e\n\u003cli\u003eMilman, A., Bonnell, C., Maguire, R., Sorensen, K., \u0026amp; Blomquist, W. (2021). Groundwater Recharge for Water Security: The Arizona Water Bank, Arizona. Case Studies in the Environment, 5(1). https://doi.org/10.1525/cse.2020.1113999\u003c/li\u003e\n\u003cli\u003eWahal, A., Mendenhall, E., \u0026amp; Giordano, M. (2023). Water in the West: Analyzing the disconnect between farmers\u0026apos; and policymakers\u0026apos; perceptions of Colorado River Basin shortages in Arizona. Environmental Science \u0026amp; Policy, 146, 57\u0026ndash;66. https://doi.org/10.1016/j.envsci.2023.05.002\u003c/li\u003e\n\u003cli\u003eRodell, M., Velicogna, I., \u0026amp; Famiglietti, J. S. (2009). Satellite-based estimates of groundwater depletion in India. Nature, 460(7258), 999-1002.\u003c/li\u003e\n\u003cli\u003eFamiglietti, J. S., Lo, M., Ho, S. L., Bethune, J., Anderson, K. J., Syed, T. H., ... \u0026amp; Rodell, M. (2011). Satellites measure recent rates of groundwater depletion in California\u0026apos;s Central Valley. Geophysical Research Letters, 38(3).\u003c/li\u003e\n\u003cli\u003eCastle, S. L., Thomas, B. F., Reager, J. T., Rodell, M., Swenson, S. C., \u0026amp; Famiglietti, J. S. (2014). Groundwater depletion during drought threatens future water security of the Colorado River Basin. Geophysical research letters, 41(16), 5904-5911.\u003c/li\u003e\n\u003cli\u003eTapley, B. D., Watkins, M. M., Flechtner, F., Reigber, C., Bettadpur, S., Rodell, M., Richter Sasgen, I., Famiglietti, J. S., Landerer, F. W., Chambers, D. P., Reager, J. T., Gardner, A. S., Save, H., Ivins, E. R., Swenson, S. C., Boening, C., Dahle, C., Wiese, D. N., Dobslaw, H., \u0026hellip; Velicogna, I. (2019). Contributions of GRACE to understanding climate change. In Nature Climate Change (Vol. 9, Issue 5, pp. 358\u0026ndash;369). Nature Publishing Group. https://doi.org/10.1038/s41558-019-0456-2 \u003c/li\u003e\n\u003cli\u003eRodell, M., \u0026amp; Reager, J. T. (2023). Water cycle science enabled by the GRACE and GRACE-FO satellite missions. Nature Water, 1(1), 47\u0026ndash;59. https://doi.org/10.1038/s44221-022-00005-0\u003c/li\u003e\n\u003cli\u003eLiu, P. W., Famiglietti, J. S., Purdy, A. J., Adams, K. H., McEvoy, A. L., Reager, J. T., ... \u0026amp; Rodell, M. (2022). Groundwater depletion in California\u0026rsquo;s Central Valley accelerates during megadrought. Nature Communications, 13(1), 7825.\u003c/li\u003e\n\u003cli\u003eRodell, M., Chen, J., Kato, H., Famiglietti, J. S., Nigro, J., \u0026amp; Wilson, C. R. (2007). Estimating groundwater storage changes in the Mississippi River basin (USA) using GRACE. Hydrogeology Journal, 15(1), 159\u0026ndash;166. https://doi.org/10.1007/s10040-006-0103-7 \u003c/li\u003e\n\u003cli\u003eRodell, M., \u0026amp; Famiglietti, J. S. (2002). The potential for satellite-based monitoring of groundwater storage changes using GRACE: the High Plains aquifer, Central US. Journal of Hydrology. www.elsevier.com/locate/jhydrol \u003c/li\u003e\n\u003cli\u003eYeh, P. J. F., Swenson, S. C., Famiglietti, J. S., \u0026amp; Rodell, M. (2006). Remote sensing of groundwater storage changes in Illinois using the Gravity Recovery and Climate Experiment (GRACE). Water Resources Research, 42(12).\u003c/li\u003e\n\u003cli\u003eAdams, K. H., Reager, J. T., Rosen, P., Wiese, D. N., Farr, T. G., Rao, S., Haines, B. J., Argus, D. F., Liu, Z., Smith, R., Famiglietti, J. S., \u0026amp; Rodell, M. (2022). Remote Sensing of Groundwater: Current Capabilities and Future Directions. In Water Resources Research (Vol. 58, Issue 10). John Wiley and Sons Inc. https://doi.org/10.1029/2022WR032219 \u003c/li\u003e\n\u003cli\u003eReager, J. T., \u0026amp; Famiglietti, J. S. (2013). Characteristic mega-basin water storage behavior using GRACE. Water Resources Research, 49(6), 3314\u0026ndash;3329. https://doi.org/10.1002/wrcr.20264 \u003c/li\u003e\n\u003cli\u003eAli, S., Liu, D., Fu, Q., Cheema, M. J. M., Pham, Q. B., Rahaman, M. M., Dang, T. D., \u0026amp; Anh, D. T. (2021). Improving the resolution of grace data for spatio-temporal groundwater storage assessment. Remote Sensing, 13. https://doi.org/10.3390/rs13173513\u003c/li\u003e\n\u003cli\u003eHumphrey, V., Gudmundsson, L., \u0026amp; Seneviratne, S. I. (2016). Assessing Global Water Storage Variability from GRACE: Trends, Seasonal Cycle, Subseasonal Anomalies and Extremes. In Surveys in Geophysics (Vol. 37, pp. 357\u0026ndash;395). Springer Netherlands. https://doi.org/10.1007/s10712-016-9367-1 \u003c/li\u003e\n\u003cli\u003eYin, W., Hu, L., Zhang, M., Wang, J., \u0026amp; Han, S. C. (2018). Statistical Downscaling of GRACE-Derived Groundwater Storage Using ET Data in the North China Plain. Journal of Geophysical Research: Atmospheres, 123(11), 5973\u0026ndash;5987. https://doi.org/10.1029/2017JD027468 \u003c/li\u003e\n\u003cli\u003eErlingis, J. M., Rodell, M., Peters-Lidard, C. D., Li, B., Kumar, S. V., Famiglietti, J. S., Granger, S. L., Hurley, J. V., Liu, P. W., \u0026amp; Mocko, D. M. (2021). A High-Resolution Land Data Assimilation System Optimized for the Western United States. Journal of the American Water Resources Association, 57, 692\u0026ndash;710. https://doi.org/10.1111/1752-1688.12910\u003c/li\u003e\n\u003cli\u003eBillah, M. M., Goodall, J. L., Narayan, U., Reager, J. T., Lakshmi, V., \u0026amp; Famiglietti, J. S. (2015). A methodology for evaluating evapotranspiration estimates at the watershed-scale using GRACE. Journal of Hydrology, 523, 574\u0026ndash;586. https://doi.org/10.1016/j.jhydrol.2015.01.066 \u003c/li\u003e\n\u003cli\u003eNiu, G.-Y., Yang, Z.-L., Mitchell, K. E., Chen, F., Ek, M. B., Barlage, M., et al. (2011). The community Noah land surface model with multiparameterization options (Noah-MP): 1. Model description and evaluation with local-scale measurements. Journal of Geophysical Research, 116(D12). https://doi.org/10.1029/2010JD015139 \u003c/li\u003e\n\u003cli\u003eHe, C., Valayamkunnath, P., Barlage, M., Chen, F., Gochis, D., Cabell, R., Schneider, T., Rasmussen, R., Niu, G. Y., Yang, Z. L., Niyogi, D., \u0026amp; Ek, M. (2023). Modernizing the open-source community Noah with multi-parameterization options (Noah-MP) land surface model (version 5.0) with enhanced modularity, interoperability, and applicability. Geoscientific Model Development, 16(17), 5131\u0026ndash;5151. https://doi.org/10.5194/gmd-16-5131-2023 \u003c/li\u003e\n\u003cli\u003eChandanpurkar, H. A., Famiglietti, J. S., Gopalan, K., Wiese, D. N., Wada, Y., Kakinuma, K., Reager, J. T., \u0026amp; Zhang, F. (2025). Unprecedented continental drying, shrinking freshwater availability, and increasing land contributions to sea level rise. Science Advances, 11(30). https://doi.org/10.1126/sciadv.adx0298 \u003c/li\u003e\n\u003cli\u003eThomas, B. F., \u0026amp; Famiglietti, J. S. (2019). Identifying climate-induced groundwater depletion in GRACE observations. Sci Rep 9: 4124. https://doi.org/10.1038/s41598-019-40155-y \u003c/li\u003e\n\u003cli\u003eLiu, P. W., Famiglietti, J. S., Purdy, A. J., Adams, K. H., McEvoy, A. L., Reager, J. T., ... \u0026amp; Rodell, M. (2022). Groundwater depletion in California\u0026rsquo;s Central Valley accelerates during megadrought. Nature Communications, 13(1), 7825. https://doi.org/10.1038/s41467-022-35582-x. \u003c/li\u003e\n\u003cli\u003eCondon, L. E., Atchley, A. L., \u0026amp; Maxwell, R. M. (2020). Evapotranspiration depletes groundwater under warming over the contiguous United States. Nature Communications, 11, 873. https://doi.org/10.1038/s41467-020-14688-0 \u003c/li\u003e\n\u003cli\u003eWu, X., Feng, X., \u0026amp; Fu, B. (2025). Climate aridification and intensified human interference undermined water storage in the Lower Yellow River region. Geography and Sustainability, 6(4), 100303. https://doi.org/10.1016/j.geosus.2025.100303\u003c/li\u003e\n\u003cli\u003eFred Tillman, B. D., Gangopadhyay, S., \u0026amp; Pruitt, T. (2020). Trends in recent historical and projected climate data for the Colorado River Basin and potential effects on groundwater availability. https://pubs.usgs.gov/publication/sir20205107 \u003c/li\u003e\n\u003cli\u003eKyl Center for Water Policy at ASU\u0026apos;s Morrison Institute. (2024). Groundwater Protection in the Valley of the Sun. https://morrisoninstitute.asu.edu/sites/default/files/2024-04/ASU-KC-GrndwtrRprt-FNL-HR-4.pdf \u003c/li\u003e\n\u003cli\u003eOhenhen, L. O., Zhai, G., Lucy, J., Werth, S., Carlson, G., Khorrami, M., Onyike, F., Sadhasivam, N., Tiwari, A., Ghobadi-Far, K., Sherpa, S. F., Lee, J.-C., Zehsaz, S., \u0026amp; Shirzaei, M. (2025). Land subsidence risk to infrastructure in US metropolises. Nature Cities, 2, 543\u0026ndash;554. https://doi.org/10.1038/s44284-025-00240-y \u003c/li\u003e\n\u003cli\u003eTadych, D. E., Ford, M., Colby, B. G., \u0026amp; Condon, L. E. (2024). Historical patterns of well drilling and groundwater depth in Arizona, considering groundwater regulation and surface water access. Journal of the American Water Resources Association, 60(6), 1193\u0026ndash;1208. https://doi.org/10.1111/1752-1688.13234 \u003c/li\u003e\n\u003cli\u003eHoerling, M. P., Eischeid, J. K., Diaz, H. F., Rajagopolan, B., \u0026amp; Kuhn, E. (2024). Critical effects of precipitation on future Colorado River flow. Journal of Climate, 37(16), 4079-4093.\u003c/li\u003e\n\u003cli\u003eArizona Department of Water Resources. (2022, September 6). ADWR\u0026apos;s review of Pinal stakeholder model revisions. https://www.azwater.gov/news/articles/2022-09-06/adwrs-review-pinal-stakeholder-model-revisions \u003c/li\u003e\n\u003cli\u003eQiu, Y., Famiglietti, J. S., Behrangi, A., Farmani, M. A., Sohi, H. Y., Gupta, A., Hung, F., Abdelmohsen, K., \u0026amp; Niu, G.-Y. (2025). The strong impact of precipitation intensity on groundwater recharge and terrestrial water storage change in Arizona, a typical dryland. Geophysical Research Letters, 52(14), e2025GL114747. https://doi.org/10.1029/2025GL114747 \u003c/li\u003e\n\u003cli\u003eMilman, A., Bonnell, C., Maguire, R., Sorensen, K., \u0026amp; Blomquist, W. (2021). Groundwater recharge for water security: the Arizona Water Bank, Arizona. Case Studies in the Environment, 5(1), 1113999.\u003c/li\u003e\n\u003cli\u003eBernat, R. F. A., Megdal, S. B., Eden, S., \u0026amp; Bakkensen, L. A. (2023). Stakeholder opinions on the issues of the Central Arizona Groundwater Replenishment District and policy alternatives. Water, 15(6), 1166. https://doi.org/10.3390/w15061166 \u003c/li\u003e\n\u003cli\u003eConway, B. D. (2015). Land subsidence and earth fissures in south-central and southern Arizona, USA. Hydrogeology Journal, 23(6), 1247\u0026ndash;1266. https://doi.org/10.1007/s10040-015-1329-z \u003c/li\u003e\n\u003cli\u003eDraper, N. R., \u0026amp; Smith, H. (1998). Applied regression analysis (3rd ed.). Wiley.\u003c/li\u003e\n\u003cli\u003eRodgers, J. L., \u0026amp; Nicewander, W. A. (1988). Thirteen ways to look at the correlation coefficient. The American Statistician, 42(1), 59\u0026ndash;66. https://doi.org/10.1080/00031305.1988.10475524 \u003c/li\u003e\n\u003cli\u003eAbdi, H., \u0026amp; Williams, L. J. (2010). Principal component analysis. In Wiley Interdisciplinary Reviews: Computational Statistics (Vol. 2, Issue 4, pp. 433\u0026ndash;459). https://doi.org/10.1002/wics.101 \u003c/li\u003e\n\u003cli\u003eArthur, D., \u0026amp; Vassilvitskii, S. (2007). k-means++: The advantages of careful seeding. In Proceedings of the eighteenth annual ACM-SIAM symposium on Discrete algorithms (pp. 1027\u0026ndash;1035). Society for Industrial and Applied Mathematics.\u003c/li\u003e\n\u003cli\u003ePedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., Vanderplas, J., Passos, A., Cournapeau, D., Brucher, M., Perrot, M., \u0026amp; Duchesnay, \u0026Eacute;. (2011). Scikit-learn: Machine learning in Python. Journal of Machine Learning Research, 12, 2825\u0026ndash;2830.\u003c/li\u003e\n\u003cli\u003eKumar, S. V., Peters-Lidard, C. D., Tian, Y., Houser, P. R., Geiger, J., Olden, S., Lighty, L., Eastman, J. L., Doty, B., Dirmeyer, P., Adams, J., Mitchell, K., Wood, E. F., \u0026amp; Sheffield, J. (2006). Land Information System\u0026mdash;An interoperable framework for high resolution land surface modeling. Environmental Modelling \u0026amp; Software, 21(10), 1402\u0026ndash;1415. https://doi.org/10.1016/j.envsoft.2005.07.004 \u003c/li\u003e\n\u003cli\u003ePeters-Lidard, C. D., Houser, P. R., Tian, Y., Kumar, S. V., Geiger, J., Olden, S., Lighty, L., Doty, B., Dirmeyer, P., Adams, J., Mitchell, K., Wood, E. F., \u0026amp; Sheffield, J. (2007). High-performance Earth system modeling with NASA/GSFC\u0026apos;s Land Information System. Innovations in Systems and Software Engineering, 3(3), 157\u0026ndash;165. https://doi.org/10.1007/s11334-007-0028-x \u003c/li\u003e\n\u003cli\u003eSohoulande, C. D., Martin, J., Szogi, A., \u0026amp; Stone, K. (2020). Climate-driven prediction of land water storage anomalies: An outlook for water resources monitoring across the conterminous United States. Journal of Hydrology, 588, 125053.\u003c/li\u003e\n\u003cli\u003eRodell, M., Houser, P. R., Jambor, U. E. A., Gottschalck, J., Mitchell, K., Meng, C. J., ... \u0026amp; Toll, D. (2004). The global land data assimilation system. Bulletin of the American \u003cem\u003eMeteorological society\u003c/em\u003e, \u003cem\u003e85\u003c/em\u003e(3), 381-394.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"GRACE satellites, WLDAS, Principal Component Analysis, Active Management Areas, Lower Colorado River Basin","lastPublishedDoi":"10.21203/rs.3.rs-7698407/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7698407/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eGroundwater resources in central and southern Arizona are increasingly threatened by prolonged drought and rising temperature, curtailed surface water supply, and pumping overdraft. Recent assessments indicate that groundwater storage in the region is rapidly declining. While human influences are well-understood, natural hydroclimatic controls on groundwater storage changes are examined in less detail. Using NASA's Gravity Recovery and Climate Experiment (GRACE) and its successor GRACE-Follow On (GRACE-FO) Groundwater Storage Anomalies (GWSA) and land-surface variables from Western Land Data Assimilation System (WLDAS), we quantified standardized trends, subbasin-scale correlations, and dominant modes via Principal Component Analysis (PCA) from 2004 to 2021 in central and southern Arizona. Using key components, we grouped groundwater subbasins using k-means clustering to simplify assessments. Results show substantial spatial heterogeneity arising from recharge-responsive northern and central subbasins linked to precipitation and subsurface runoff, and southern loss-dominated subbasins linked to weaker recharge coupling and more substantial influence from atmospheric demand. PCA analysis shows that natural modes variability explains about 16% of the spatial variance displayed in GRACE/FO GWSA in the study area, with approximately 72% of that variance explained by total evapotranspiration (\u0026asymp;\u0026thinsp;29%), precipitation (\u0026asymp;\u0026thinsp;23%), and subsurface runoff (\u0026asymp;\u0026thinsp;20%). The remaining variance reflects a combination of anthropogenic influences (pumping, MAR), geologic heterogeneity, and residual model/measurement errors. Our diagnostic framework presents groundwater subbasin scale clusters driven by shared hydroclimatic modes of variability. It has potential as a transferable tool for focused recharge, groundwater sustainability plans, and post-2026 groundwater planning in the Lower Colorado River Basin.\u003c/p\u003e","manuscriptTitle":"Key natural influences on groundwater storage changes in central and southern Arizona","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-17 10:20:42","doi":"10.21203/rs.3.rs-7698407/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-01-02T19:00:48+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-11-30T19:29:26+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"93949106309888419696833362633809794097","date":"2025-11-01T20:03:59+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-10-31T18:35:58+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"91621945632799494282107840018151438558","date":"2025-10-20T19:23:04+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-10-06T14:15:52+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-10-06T14:04:02+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-10-06T08:21:58+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-09-26T22:36:51+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-09-26T22:33:30+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"ffde495f-7096-40ba-a2e5-dd6f710b5d10","owner":[],"postedDate":"October 17th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":56419173,"name":"Earth and environmental sciences/Environmental sciences"},{"id":56419174,"name":"Earth and environmental sciences/Hydrology"},{"id":56419175,"name":"Earth and environmental sciences/Solid earth sciences"}],"tags":[],"updatedAt":"2026-03-30T16:32:50+00:00","versionOfRecord":{"articleIdentity":"rs-7698407","link":"https://doi.org/10.1038/s41598-026-44132-0","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2026-03-24 16:10:08","publishedOnDateReadable":"March 24th, 2026"},"versionCreatedAt":"2025-10-17 10:20:42","video":"","vorDoi":"10.1038/s41598-026-44132-0","vorDoiUrl":"https://doi.org/10.1038/s41598-026-44132-0","workflowStages":[]},"version":"v1","identity":"rs-7698407","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7698407","identity":"rs-7698407","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}