The Integrated Multisatellite Retrievals for Global Precipitation Measurement (GPM) (IMERG) Level 3 estimates rainfall from passive microwave sensors onboard satellites that are associated with several uncertainty sources such as sensor calibration, retrieval errors, and orographic effects. This study aims to provide a comprehensive investigation of multiple machine learning (ML) techniques (Random Forest, and Neural Networks), to stochastically generate an error-corrected improved IMERG precipitation product at a daily time scale and 0.1°-degree spatial resolution over the Brahmaputra river basin. In this study, we used the operational IMERG-Late Run version 06 product along with several meteorological and land surface parameters (elevation, soil type, land type, soil moisture, and daily maximum and minimum temperature) to produce an improved precipitation product in the Brahmaputra basin. We trained, tested, and optimized ML algorithms using 4 years (from 2015 through 2019) of reference rainfall data derived from the rain gauge. The ML generated precipitation product exhibited improved systematic and random error statistics for the study area, which is a strong indication for using the proposed algorithms in retrieving precipitation across the globe. We conclude that the proposed ML-based ensemble framework has the potential to quantify and correct the error sources for improving and promoting the use of satellite-based precipitation estimates for water resources applications.

The evaluation of water systems based on historical statistics is problematic when shifts in the hydrologic system occur due to a changing climate. An explicit link to thermodynamic and dynamic pathways in the climate system that may control water system performance is missing from current operational policies prescribed by regulation manuals within the US. In response, this study contributes an extended version of an existing weather regime (WR)–based stochastic weather generator (SWG) that allows (1) hourly simulation, (2) over the entire year, and (3) with a corrected representation of extremes. A range of climate scenarios is developed to demonstrate the insights that can be gained from linking the impacts of climate change to their thermodynamic and dynamic causal mechanisms, in this case for inflows to the Don Pedro Reservoir within the Tuolumne River Watershed of California. Application of the WR-SWG and water system modeling chain shows that the magnitude of flood events can be heavily influenced by antecedent hydrologic factors such as snow water equivalent (SWE) and soil moisture. Our results suggest that, under all climate change scenarios, SWE decreases as temperature increases and contributes more (sometimes up to 2.5 times more than the baseline) inflow as part of rain-on-snow events. The monthly reservoir inflows show the potential to cause extreme floods as the average rate of inflow increases by up to 80% with temperature increases, whereas SWE tends to increase by 50%, adding water to the stream during the high flow season. In addition to the temperature increase, if the water-holding capacity of the atmosphere increases with Clausius-Clapeyron scaling, reservoir inflows are projected to increase. This provides insight for risk-hedging policies: winter storm and spring snowmelt release and storage decisions that drive flood and drought risk, respectively.

The threat that climate change poses to water resource systems has led to a substantial and growing number of impact studies. These studies follow two approaches: (a) top‐down studies are driven by projections of future climate change provided by downscaled general circulation models (GCMs); and (b) bottom‐up studies are driven by the systematic evaluation of exploratory scenarios. Top‐down approaches produce realistic scenarios rooted in the simulation of thermodynamic and dynamic processes represented in GCMs, but the internal resolution of these processes make it difficult to link vulnerabilities to discrete components of change. Bottom‐up approaches link vulnerabilities to discrete components of change through the structured evaluation of exploratory scenarios, but the lack of insight rooted in climate change processes can lead to the development of implausible scenarios. This paper evaluates exploratory scenarios developed through thermodynamic and dynamical guided perturbations motivated by GCM‐bound insights. The result is a hybrid approach that bridges a gap between top‐down and bottom‐up approaches. This yields several advantages. First, emerging vulnerabilities are linked to distinct thermodynamic and dynamic processes that are modeled in GCMs with differential likelihoods and plausible ranges of change. Second, the structured evaluation of process‐informed exploratory scenarios link system vulnerabilities to distinct components of climate change. An illustrative case study demonstrates perturbations linked to thermodynamic and dynamical processes have a large impact on stakeholder‐defined flood and drought performance, and the structured evaluation of process‐informed exploratory scenarios find nuanced infrastructure‐specific vulnerabilities that would be difficult to identify using an alternative approach.

The threat that climate change poses to water resource systems has led to a significant and growing number of impact studies. These studies tend to follow two general methodological approaches: (1) top-down, process-based studies driven by projections of future climate change supplied by downscaled general circulation models (GCMs), and (2) bottom-up, vulnerability-based studies driven by exploratory scenarios. Top-down studies generate realistic climate scenarios, but computational burdens limit the ensemble size. As a result, critical vulnerabilities may be left unexplored. Bottom-up approaches make it possible to assess a wide range of scenarios, but usually without connection to physically plausible climate processes, limiting their utility in adaptive planning. This study develops process-informed exploratory scenarios that bridge the gap between top-down and bottom-up methods. This hybrid approach yields several advantages. First, emerging vulnerabilities associated with non-linear hydrologic changes are linked to thermodynamic and dynamic climate drivers modeled in the GCMs with differential likelihoods and plausible ranges of change. This provides a transparent link between stakeholder defined vulnerabilities and climate processes that is often missing in bottom-up assessments. Second, non-linear shifts in vulnerability are directly linked to specific climate drivers, through the systematic perturbation of process informed climate variables. Making this connection in top-down assessments is difficult since the climate response to an emissions scenario is modeled as part of an endogenous process. The hybrid approach developed by this study is presented with a case study in the Tuolumne River watershed; through which thermodynamic and dyanamically guided climate scenarios were created by a process-informed stochastic weather generator to evaluate flood and drought related performance vulnerabilities at the New Don Pedro Dam near the watershed’s outlet. This case study finds that flood and drought performance at the dam is more sensitive to process-informed climate drivers than less theoretically grounded delta shifts precipitation, and non-linear system responses to climate drivers are revealed through the systematic perturbation process-informed climate variables.