Search

Advances in satellite Earth observation have opened up new opportunities for a global monitoring of soil moisture (SM) at fine to medium resolution, but satellite remote sensing can only measure the near-surface soil moisture (SSM). As such, it is critically important to examine the potential of satellite SSM measurements to derive the water resource variations 15 in deeper subsurface. This study compares the SSM variability captured by the Soil Moisture Active and Passive (SMAP) satellite and the Soil Water Index (SWI) derived from SMAP SSM with subsurface SM and groundwater (GW) dynamics simulated by a high resolution fully-integrated surface water - groundwater model over an agriculturally-dominated watershed in eastern Canada across two spatial scales, namely SMAP product grid (9 km) and watershed (-4000 km2). SMAP measurements compare well with the hydrologic simulations in terms of SSM variability at both scales. Simulated 20 subsurface SM and GW storage show lagged and smoother characteristics relative to SMAP SSM variability with an optimal delay of -1 days for the 25-50 cm SM, -6 days for the 50-100 cm SM, and -11 days for the GW storage for both scales. Modelled subsurface SM dynamics agree well with the SWI derived from SMAP SSM using the classic characteristic time lengths (15 days for the 0-25 cm layer and 20 days for the 0-100 cm layer). The simulated GW storage showed a slightly delayed variation relative to the derived SWI. The quantified optimal characteristic time length Topt for SWI estimation (by 25 matching the variations in SMAP-derived SWI and modeled root zone SM) is comparable to Topt obtained in other agricultural regions around the world. This work demonstrates SMAP SM measurements as a potentially useful aid when predicting root zone SM and GW dynamics and validating fully integrated hydrologic models across different spatial scales. This study also provides insights into the dynamics of near surface--subsurface water interaction and the capabilities and approaches of satellite-based SM monitoring and high resolution fully-integrated hydrologic modelling. [ABSTRACT FROM AUTHOR]

Water is essential for all ecosystem services, yet a comprehensive assessment of total (overall) water contributions to ecosystem services production has never been attempted. Quantification of the many ecosystem services impacted by water demands integrated hydrological simulations that implicitly characterize subsurface and surface water exchange. In this study, we use a fully integrated hydrological model—HydroGeoSphere (HGS)—to capture changes in subsurface water, surface water, and evapotranspiration (green water) combined with the economic valuation approach to assess ecosystem services over an 18-year period (2000–2017) in a mixed-use but predominantly agricultural watershed in eastern Ontario, Canada. Using the green water volumes and ecosystem services values as inputs, we calculate the marginal productivity of water, which is $0.45 per m3 (in 2022 Canadian dollars). The valuation results show that maximum green water is used during the dry years, with a value of $1.16 billion during a severe drought that struck in 2012. The average product of water for ecosystem services declines during the dry years. Because subsurface water is a major contributor to the green water supply, it plays a critical role in sustaining ecosystem services during drought conditions. For instance, during the 2012 drought, the subsurface water contribution to green water was estimated at $743 million, making up 72 % of the total value of green water used in that year. Conversely, the surface water contributions in green water provision over the modeling period are comparatively miniscule. This study informs watershed management on the sustainable use of subsurface water during droughts and provides an improved methodology for watershed-based integrated management of ecosystem services. [ABSTRACT FROM AUTHOR]

Snow is a critical contributor to Ontario's water-energy budget with impacts on water resource management and flood forecasting. Snow water equivalent (SWE) describes the amount of water stored in a snowpack and is important in deriving estimates of snowmelt. However, only a limited number of sparsely distributed snow survey sites (n = 383) exist throughout Ontario. The SNOw Data Assimilation System (SNODAS) is a daily, 1 km gridded SWE product that provides uniform spatial coverage across this region; however, we show here that SWE estimates from SNODAS display a strong positive mean bias of 50 % (16 mm SWE) when compared to in situ observations from 2011 to 2018. This study evaluates multiple statistical techniques of varying complexity, including simple subtraction, linear regression and machine learning methods to bias-correct SNODAS SWE estimates using absolute mean bias and RMSE as evaluation criteria. Results show that the random forest (RF) algorithm is most effective at reducing bias in SNODAS SWE, with an absolute mean bias of 0.2 mm and RMSE of 3.64 mm when compared with in situ observations. Other methods, such as mean bias subtraction and linear regression, are somewhat effective at bias reduction however, only the RF method captures the nonlinearity in the bias, and its interannual variability. Applying the RF model to the full spatio-temporal domain shows that the SWE bias is largest before 2015, during the spring melt period, north of 44.5° N and East (downwind) of the Great Lakes. As an independent validation, we also compare estimated snowmelt volumes with observed hydrographs, and demonstrate that uncorrected SNODAS SWE is associated with unrealistically large volumes at the time of the spring freshet, while bias-corrected SWE values are highly consistent with observed discharge volumes. [ABSTRACT FROM AUTHOR]

This study aims to assess the impact of climate change on water resources in a large watershed within the Laurentian Great Lakes region, using the fully integrated surface‐subsurface model HydroGeoSphere. The hydrologic model is forced with an ensemble of high‐resolution climate projections from the Weather Research and Forecasting (WRF) model. The latter has been extended with an interactive lake model (FLake) to capture the effect of the Great Lakes on the regional climate. The WRF ensemble encompasses two different moist physics configurations at resolutions of 90, 30, and 10 km, as well as four different initial and boundary conditions, so as to control for natural climate variability. The integrated hydrologic model is run with a representative seasonal cycle, which effectively controls natural climate variability, while remaining computationally tractable with a large integrated model. However, the range of natural variability is also investigated, as are the impacts of climate model resolution and bias correction. The two WRF configurations show opposite climate change responses in summer precipitation, but similar responses otherwise. The hydrologic simulations generally follow the climate forcing; however, due to the memory of the subsurface, the differences in summer propagate throughout the entire seasonal cycle. This results in a set of dry scenarios with reduced streamflow and water availability year‐round and a set of wet scenarios with increased streamflow for all times excluding the spring peak, which does not increase. Most of the analysis focuses on streamflow, but changes in the seasonal cycle of baseflow and groundwater recharge are also analyzed. Plain Language Summary: In this study we investigate the impact of climate change on water resources using state‐of‐the‐art computer simulations. The simulations were conducted using physically based models, which simulate the circulation of the atmosphere, rainfall, and the flow of water above and below the surface. The region of interest here is the Grand River Watershed, located in the Great Lakes region (southern Ontario, Canada). We show that with high resolution and a physical representation of the Great Lakes, only a simple correction—much less than for global climate models—is necessary, in order to simulate a realistic climate. We find that predicted climate change impacts on water resources depend strongly on some approximations commonly made to represent thunderstorms and precipitation in climate models. The most likely scenario, based on our analysis, is an increase in precipitation and streamflow in all seasons except spring, but some scenarios also show less precipitation in summer, which results in lower streamflow year‐round. A major result is that differences in summer precipitation can affect streamflow in all seasons, but only if the interaction with groundwater is properly accounted for. At the moment uncertainty in future summer precipitation changes limits our ability to predict impacts on water resources. Key Points: The effective use of an RCM ensemble for integrated hydrologic modeling using a representative average seasonal cycle is exploredWithin the study area, the largest uncertainty in hydrological modeling of climate change impacts is the change in summer precipitationChanges in summer precipitation affect groundwater storage, which can significantly impact baseflow throughout the year [ABSTRACT FROM AUTHOR]