Search

Recently, more advanced synchronous global‐scale satellite observations, the Soil Moisture Active Passive enhanced Level 3 (SMAP L3) soil moisture product and the Orbiting Carbon Observatory 2 (OCO‐2) solar‐induced chlorophyll fluorescence (SIF) product, provide an opportunity to improve the predictive understanding of both water and carbon cycles in land surface modeling. The Simplified Simple Biosphere Model version 4 (SSiB4) was coupled with the Top‐down Representation of Interactive Foliage and Flora Including Dynamics Model (TRIFFID) and a mechanistic representation of SIF. Incorporating dynamic vegetation processes reduced global SIF root‐mean‐squared error (RMSE) by 12%. Offline experiments were conducted to understand the water and carbon cycles and their interactions using satellite data as constraints. Results indicate that soil hydraulic properties, the soil hydraulic conductivity at saturation (Ks) and the water retention curve, significantly impact soil moisture and SIF simulation, especially in the semi‐arid regions. The wilting point and maximum Rubisco carboxylation rate (Vmax) affect photosynthesis and transpiration, then soil moisture. However, without atmospheric feedback processes, their effects on soil moisture are undermined due to the compensation between soil evaporation and transpiration. With optimized parameters based on SMAP L3 and OCO‐2 data, the global RMSE of soil moisture and SIF simulations decreased by 15% and 12%, respectively. These findings highlight the importance of integrating advanced satellite data and dynamic vegetation processes to improve land surface models, enhancing understanding of terrestrial water and carbon cycles. This study used advanced satellite observations and land surface models to learn more about the water and carbon cycles and their interactions. The soil hydraulic conductivity at saturation (Ks) and the parameter for the water retention curve were identified as important. They affect how fast water can move in soil and how much water can be held by soil, which affects the available water in the soil for plants to use and makes them particularly important in water and carbon cycle simulation in dry areas. Two vegetation parameters affecting photosynthesis can help to improve the carbon cycle simulation but cannot change the soil moisture much because of the offline model limitation. The adjustment on the two soil property parameters and the two vegetation parameters improved the global water cycle simulation by 15% and the carbon cycle simulation by 12%. The introduction of dynamic vegetation processes into our model improved the carbon cycle simulation by 12% globally and helped to better understand the water and carbon cycles and their interactions. Key processes and parameters are identified using satellite data for advances in understanding carbon and water cycles and interactionsThe soil hydraulic properties, wilting point, and maximum Rubisco carboxylation rate are key parameters linking water‐carbon cyclesThe dynamic vegetation process plays an important role in water and carbon cycles Key processes and parameters are identified using satellite data for advances in understanding carbon and water cycles and interactions The soil hydraulic properties, wilting point, and maximum Rubisco carboxylation rate are key parameters linking water‐carbon cycles The dynamic vegetation process plays an important role in water and carbon cycles

Drought is an understudied driver of infectious disease dynamics. Amidst the ongoing southwestern North American megadrought, California (USA) is having the driest multi-decadal period since 800 CE, exacerbated by anthropogenic warming. In this study, we aimed to examine the influence of drought on coccidioidomycosis, an emerging infectious disease in southwestern USA.

The California state government put restrictions on outdoor residential water use, including landscape irrigation, during the 2012–2016 drought. The public health implications of these actions are largely unknown, particularly with respect to mosquito-borne disease transmission. While residential irrigation facilitates persistence of mosquitoes by increasing the availability of standing water, few studies have investigated its effects on vector abundance. In two study sub-regions in the Los Angeles Basin, we examined the effect of outdoor residential water use restrictions on the abundance of the most important regional West Nile virus vector, Culex quinquefasciatus. Using spatiotemporal random forest models fit to Cx.abundance during drought and non-drought years, we generated counterfactual estimates of Cx.abundance under a hypothetical drought scenario without water use restrictions. We estimate that Cx.abundance would have been 44% and 39% larger in West Los Angeles and Orange counties, respectively, if outdoor water usage had remained unchanged. Our results suggest that drought, without mandatory water use restrictions, may counterintuitively increase the availability of larval habitats for vectors in naturally dry, highly irrigated settings and such mandatory water use restrictions may constrain Cx.abundance, which could reduce the risk of mosquito-borne disease while helping urban utilities maintain adequate water supplies.

Ongoing runoff declines in the Colorado River Basin have been shown to be predominately driven by decreasing albedo from warming‐driven snow‐cover loss, especially in late‐spring (hereafter snowmelt‐radiation feedback). Here, we explore the feedback's impact on annual runoff sensitivity to warming across the western U.S. (WUS) using hydrologic model simulations. For 1°C uniform warming, we show that runoff is most sensitive to warming in modestly snow‐covered, interior mountain headwaters, especially the Rocky Mountains. Runoff sensitivities are most associated with the snowmelt‐radiation feedback in basins with runoff coefficients between 0.2 and 0.6, where runoff sensitivity increases with more snow and lower winter temperature. In aggregate, ∼48% of WUS runoff sensitivity is attributable to the snowmelt‐radiation feedback and is especially pronounced in the warming‐sensitive river basins (annual runoff decreases >5%/°C). We also show that the feedback's impact decreases with increasing temperature, which has unresolved implications for streamflow declines in a less‐snow future. Regional climate warming is driving strong runoff changes in the western U.S. (WUS), especially the Upper Colorado River Basin (UCRB). Previous work showed that warming‐related snow cover reductions lead to more solar radiation absorption and evapotranspiration, which largely explain ongoing runoff declines in UCRB. Here, we assess the impact of this snowmelt‐radiation feedback on warming‐induced runoff changes across WUS. In a warmer world, we find that the largest annual runoff sensitivities are in the interior mountainous WUS with modest snow cover. The snowmelt‐radiation feedback explains over half of the warming‐induced runoff changes in warming‐sensitive WUS basins and about half of WUS' overall runoff sensitivity. In areas influenced by the snowmelt‐radiation feedback, both runoff sensitivity and the feedback's contribution become smaller with higher temperatures, suggesting a potentially slower rate of streamflow decline as temperatures rise in a warmer future. Snowmelt‐radiation feedback accounts for ∼1/2 of warming‐driven runoff decline across the Western U.S. (WUS)Runoff sensitivities are most linked to snowmelt‐radiation feedback in river basins with runoff coefficients in the range 0.2–0.6Runoff sensitivities to warming are largest in modestly snow‐covered, interior mountainous parts of WUS, especially the Rocky Mountains Snowmelt‐radiation feedback accounts for ∼1/2 of warming‐driven runoff decline across the Western U.S. (WUS) Runoff sensitivities are most linked to snowmelt‐radiation feedback in river basins with runoff coefficients in the range 0.2–0.6 Runoff sensitivities to warming are largest in modestly snow‐covered, interior mountainous parts of WUS, especially the Rocky Mountains

Based on a process‐level modeling of the rain‐on‐snow (ROS) events in the period of 1950 to 2013 and in a warmer climate, we quantify the historical and future runoff contribution from ROS to extreme floods and the source of runoff and snowmelt in large ROS events within the conterminous United States (CONUS). We find that the regions impacted most heavily by ROS include the West Coast, the major mountain ranges of the western interior, the Upper Midwest, the Northeast, and the lower Appalachians. While 70% of extreme (upper 0.1%) runoff events in these regions have some contribution from ROS, the runoff generated during these ROS events accounts for less than 10% of the total extreme flood runoff; the much larger fraction of extreme runoff is from either intense rainfall or clear‐sky snowmelt. Rainfall is the dominant source of runoff in ROS events along the West Coast and over the west‐facing slopes of the Cascades and Sierra Nevada, while snowmelt dominates ROS runoff in the other regions in the CONUS. Net radiation dominates the snowmelt during ROS in the high mountains in the West, while net radiation and turbulent heat flux are equally dominant in the rest of CONUS. Historically, the role of ROS in streamflow extremes is most significant in midelevation areas, but this “significant influence zone” will shift to higher elevations in a warmer future. The future ROS frequency changes exert a first order control on the future change of the runoff contribution from ROS to extreme floods. Rain‐on‐snow (ROS) occurs during periods when liquid precipitation falls on a pre‐existing snowpack. The intense rainfall and the rapid snowmelt during large ROS events have a long history of contributing to floods that lead to costly property damages and even the loss of life. For example, the 2017 Oroville Dam crisis that caused damages exceeding $1B was caused in part by ROS flooding. Through an analysis of long‐term ROS and flood records over the conterminous United States, we evaluate the contribution of ROS to past flooding, and how ROS's contribution to floods will change in a warmer future. Historically, ROS has contributed to many historical flood events across the conterminous United States, especially at low to intermediate elevations (between 1,000m and 1,500 m). However, its overall contribution to the total flood runoff mostly is modest. Radiative energy, primarily the longwave radiation emitted from the moist atmosphere, collectively explains the most snowmelt during ROS. In a warmer future, the contribution of ROS to floods runoff will generally increase at high elevations (above 2,000 m) in the Western United States and will mostly decrease in the rest of the conterminous United States. Rain on snow affects many flood events across the conterminous United States, but its contribution to total flood runoff mostly is modestRain‐on‐snow frequency exerts a first‐order control on the role of rain on snow in hydrologic extremesNet radiation is overall the most dominant energy source for the snowmelt during rain on snow in most of the conterminous United States

Snow water equivalent (SWE), particularly in mountains regions, has been an elusive hydrologic measurement. We examine the utility of a data assimilation approach to generate space‐time continuous estimates of SWE from more readily available snow depth (SD) measurements. A multitemporal lidar data set provides a unique opportunity to assimilate single SD images and verify posterior estimates against SD images at nonassimilation times. Application over three water years shows significant improvement in the posterior estimates with an average correlation between estimated and measured SD fields of 0.88 compared to 0.52 for prior estimates that do not benefit from the assimilated SD data. We also show that posterior estimates are consistent with independent in situ SWE and streamflow measurements. This work demonstrates that using high‐resolution/high‐accuracy, but infrequent, SD measurements combined with a data assimilation framework could make significant inroads toward the goal of spatially distributed SWE and snowmelt estimates at the global scale. The mass of water stored in seasonal snowpacks (snow water equivalent) is a key parameter for water supply management. Yet, despite its importance and decades of effort, snow water equivalent has been an elusive variable to characterize over the mountainous areas of the globe, with no existing satellite capable of providing routine estimates of these important water stores. We propose a new approach using so‐called data assimilation techniques to transform infrequently available measurements of snow depth to continuous estimates of snow water equivalent and snowmelt rates. We show that even a single snow depth measurement near the time of peak snow accumulation can result in accurate estimation of the evolution of snow water equivalent and snowmelt rates across the water year. Our results suggest a potential new pathway for the hydrologic community in its long‐pursued goal of global‐scale estimates (from satellite remote sensing) of snow water equivalent and its space‐time variability. If applied at large scales with satellite‐derived snow depth measurements, such methods could have significant implications for improved water resource and hazard management. Data assimilation is proposed as an ideal tool for estimating continuous snow water equivalent from infrequent snow depth measurementsEven a single midseason snow depth image can provide significant information on snow water equivalent evolution through the snow seasonSnow depth data assimilation suggests an important pathway for future satellite missions aimed at retrieving snow water equivalent

River discharge estimation requires knowledge of bathymetry. However, aside from a few locations where surveys have been conducted, bathymetric data are unavailable, even for major rivers. It has been suggested that water surface elevation and flow width measurements from the upcoming Surface Water and Ocean Topography (SWOT) satellite mission (planned launch 2021) may be used to infer the submerged channel geometry; however, the full potential of these measurements for inferring bathymetry has yet to be explored. We apply four different techniques, with varying assumptions about height‐width relationships, to predict unknown bathymetry. We call these “curve‐fitting methods” the linear, slope break, nonlinear, and nonlinear slope break (NLSB) methods. The linear and slope break methods are based on a linear height‐width relationship, while the nonlinear and NLSB methods are based on a height‐width relationship derived from hydraulic geometry equations. We generate SWOT‐like observations of height and width based on 5‐m gridded Upper Mississippi River data and evaluate the performance of each curve‐fitting method given the SWOT‐like observations. The NLSB method predicts bed elevation and low flow area with the least error, although the nonlinear method may be preferred in low data conditions. Additionally, we show that our method outperforms previously suggested methods, and we propose an NLSB‐based bathymetry prior for Bayesian discharge estimation algorithms. We predict unobserved bathymetry using a limited number of error‐corrupted height and width measurementsWe compare four different height‐width models and find that piecewise, nonlinear models predict bathymetry with the least errorWe develop bathymetry priors for discharge estimation algorithms that rely on prior distributions to constrain unknown parameters

We examine the relative contributions of root zone soil moisture (SM) and groundwater (GW) storage to trends and interannual variability in total water storage (TWS) from essentially the entire Gravity Recovery and Climate Experiment (GRACE) record (2003–2016). Our study region is the Eastern United States where GW generally is shallow, and we use observations and simulations from a suite of land surface models (LSMs) where observations are not available. We first consider Illinois, which has a long‐term network of SM sensors and observation wells. We assessed the uncertainties in GRACE TWS in comparison with the observation‐based TWS using observed SM, well‐derived GW, and model‐reconstructed snow water equivalent. We evaluated LSM SM compared with observations over Illinois; generally good agreement encouraged us to use the LSM‐ensemble average SM, well‐derived GW, and Snow Data Assimilation System snow water equivalent over the entire Eastern United States to examine the relative contribution of storage components to GRACE TWS over this larger domain. Taken together, the three components explained 2%–85% of interannual variability in GRACE TWS across the 2‐digit Hydrologic Unit Code regions. Root zone SM was the dominant contributor to the interannual variability (but not trends) of GRACE TWS over much of the Eastern United States. We also compared three independent approaches to estimating GW: well data, baseflow observations, and GRACE. Despite much smaller magnitudes of variation, baseflow‐derived GW correlated more highly with well‐derived GW over the eastern‐most and Upper Mississippi regions, while GRACE‐derived GW matched better with well‐derived GW over the remaining regions. Over the Eastern United States, soil moisture accounts for most of the interannual variability (but not trends) of terrestrial water storageModels that do not represent groundwater underestimate interannual variations in terrestrial water storage by about one third

Millions of people died due to famines in India in the nineteenth and twentieth centuries; however, the relationship of historical famines with drought is complicated and not well understood. Using station‐based observations and simulations, we reconstruct soil moisture (agricultural) drought in India for the period 1870–2016. We show that over this century and a half period, India experienced seven major drought periods (1876–1882, 1895–1900, 1908–1924, 1937–1945, 1982–1990, 1997–2004, and 2011–2015) based on severity‐area‐duration analysis of reconstructed soil moisture. Out of six major famines (1873–74, 1876, 1877, 1896–97, 1899, and 1943) that occurred during 1870–2016, five are linked to soil moisture drought, and one (1943) was not. The three most deadly droughts (1877, 1896, and 1899) were linked with the positive phase of El Niño–Southern Oscillation. Five major droughts were not linked with famine, and three of those five nonfamine droughts occurred after Indian independence in 1947. India witnessed some of the most famous famines during the late nineteenth and early twentieth centuries. These famines caused millions of deaths primarily due to widespread crop failure. However, the role of agricultural drought in these famines remains unrecognized. Using station‐based observations and simulations from a hydrological model, we reconstructed agricultural droughts and established a linkage between famines and droughts over India. We find that a majority of famines were caused by large‐scale and severe soil moisture droughts that hampered the food production. However, one famine was completely due to the failure of policy during the British era. Expansion of irrigation, better public distribution system, rural employment, and transportation reduced the impact of drought on the lives of people after the independence. Major drought events have been reconstructed in India during 1870–2016India experienced seven major drought periods (1876–1882, 1895–1900, 1908–1924, 1937–1945, 1982–1990, 1997–2004, and 2011–2015)Out of six major famines that occurred during 1870–2016, five are linked to soil moisture drought, and one (1943) was not

We characterize the sensitivity of atmospheric river (AR)‐derived seasonal snowfall estimates to their atmospheric reanalysis‐based detection over Sierra Nevada, USA. We use an independent snow data set and the ARs identified with a single detection method applied to multiple atmospheric reanalyses of varying horizontal resolutions, to evaluate orographic relationships and contributions of individual ARs to the seasonal cumulative snowfall (CS). Spatial resolution differences have relatively minor effects on the number of ARs diagnosed, with higher‐resolution data sets identifying four more AR days per year, on average, during the 1985–2015 winters. However, this can lead to ~10% difference in AR attribution to the mean domain‐wide seasonal CS and differences up to 47% snowfall attribution at the seasonal scale. We show that identifying snow‐bearing ARs provides more information about the seasonal CS than simply knowing how many ARs occurred. Overall, we find that higher‐resolution atmospheric reanalyses imply greater attribution of seasonal CS to ARs. While it is known that elongated moisture‐rich atmospheric features, known as atmospheric rivers (ARs), play an important role in the water resources of the mountainous western United States, less is known about how the atmospheric data sets used to diagnose the presence of ARs influence the amount of AR‐attributed snowfall estimated each winter. Nonetheless, this missing information can be important for managing water resources and improving seasonal snowfall forecasts for areas depending on AR‐derived snowfall. We show that using a single AR detection algorithm, applied to multiple atmospheric reanalyses to identify ARs, higher‐resolution atmospheric reanalyses diagnose up to four more ARs per winter implying 10% greater AR attribution to the mean seasonal snowfall across Sierra Nevada, USA. Understanding how different atmospheric reanalyses play a role in our interpretation of AR impacts for hydrologic studies and water resources management, as we investigate here, is important especially as more ARs are projected to occur in a warmer future atmosphere. Atmospheric rivers (ARs) diagnosed with four atmospheric reanalyses often yield consistent mean seasonal snowfall patterns in the Sierra NevadaHigher‐resolution atmospheric reanalyses diagnose up to four more ARs per winter implying 10% greater AR attribution to the mean seasonal snowfallKnowing how many snowy ARs occur per winter yields more information about the total seasonal snowfall than only knowing how many ARs occur

Despite evidence of increasing precipitation extremes, corresponding evidence for increases in flooding remains elusive. If anything, flood magnitudes are decreasing despite widespread claims by the climate community that if precipitation extremes increase, floods must also. In this commentary we suggest reasons why increases in extreme rainfall are not resulting in corresponding increases in flooding. Among the possible mechanisms responsible, we identify decreases in antecedent soil moisture, decreasing storm extent, and decreases in snowmelt. We argue that understanding the link between changes in precipitation and changes in flooding is a grand challenge for the hydrologic community and is deserving of increased attention. It is now well established that rising temperatures are increasing precipitation extremes. This has led many to believe that flood magnitude and hence risk are also increasing, while observational evidence suggests otherwise. This commentary outlines the reasons for this dichotomy and presents mechanisms that may be contributing to it. The implications of increasing precipitation extremes leading to reducing flood magnitudes are discussed, and an argument is made that understanding this changing link between the two is deserving of increased attention. Extreme precipitation is increasing with rising temperaturesFlood magnitudes, however, are decreasing at the same timeHowever, this is not a complete story; very rare floods are rising while frequent floods are reducing; reasons for this are explored

The Pacific Northwest is the most highly glacierized region in the conterminous United States (858 glaciers; 466 km2). These glaciers have displayed ubiquitous patterns of retreat since the 1980s mostly in response to warming air temperatures. Glacier melt provides water for downstream uses including agricultural water supply, hydroelectric power generation, and for ecological systems adapted to cold reliable streamflow. While changes in glacier area have been studied within the region over an extended period of time, the hydrologic consequences of these changes are not well defined. We applied a high‐resolution glacio‐hydrological model to predict glacier mass balance, glacier area, and river discharge for the period 1960–2099. Six river basins across the region were modeled to characterize the regional hydrological response to glacier change. Using these results, we generalized past and future glacier area change and discharge across the entire Pacific Northwest using a k‐means cluster analysis. Results show that the rate of regional glacier recession will increase, but the runoff from glacier melt and its relative contribution to streamflow display both positive and negative trends. In high‐elevation river basins enhanced glacier melt will buffer strong declines in seasonal snowpack and decreased late summer streamflow, before the glaciers become too small to support streamflow at historic levels later in the 21st century. Conversely, in lower‐elevation basins, smaller snowpack and the shrinkage of small glaciers result in continued reductions in summer streamflow. Spatial and temporal patterns of glacier retreat and hydrological response are illustrated for the Pacific Northwest United States 1960–2099The relative contribution of glacier melt in space and time is characterized at the basin and channel network scaleThe response of summer streamflow to glacier recession varies greatly with respect to local climate

The Colorado River is the primary surface water resource in the rapidly growing U.S. Southwest. Over the period 1916–2014, the Upper Colorado River Basin naturalized streamflow declined by 16.5%, despite the fact that annual precipitation in the UCRB over that period increased slightly (+1.4%). In order to examine the causes of the runoff declines, we performed a set of experiments with the Variable Infiltration Capacity hydrology model. Our results show that the pervasive warming has reduced snowpacks and enhanced evapotranspiration over the last 100 years; over half (53%) of the long‐term decreasing runoff trend is associated with the general warming. Negative winter precipitation trends have occurred in the handful of highly productive subbasins that account for over half of the streamflow at Lee's Ferry. We also compared a midcentury drought with the (ongoing) post‐Millennium Drought and find that whereas the earlier drought was caused primarily by pervasive low‐precipitation anomalies across UCRB, higher temperatures have played a large role in the post‐Millennium Drought. The post‐Millennium Drought has also been exacerbated by negative precipitation anomalies in several of the most productive headwater basins. Finally, we evaluate the UCRB April–July runoff forecast for 2017, which decreased dramatically as the runoff season progressed. We find that while late winter and spring 2017 was anomalously warm, the proximate cause of most of the forecast reduction was anomalous late winter and early spring dryness in UCRB, which followed exceptionally large (positive) early winter precipitation anomalies. As the essential water resource for the Southwest United States, the Upper Colorado River Basin (UCRB) unimpaired streamflow declined by 16.5% over 1916–2014, while annual precipitation increased slightly (+1.4%). We performed a set of experiments with a hydrology model that uses temperature and precipitation as inputs to diagnose the causes of this apparent anomaly. We find that over half (53%) of the decreasing runoff trend is associated with unprecedented basin‐wide warming, which has reduced snowpack and increased plant water use. The remaining ~47% of the trend is associated mostly with reduced winter precipitation in four highly productive subbasins, all located in Colorado. We compared the 1953–1967 drought with the 2000–2014 Millennium Drought and find that the earlier drought was caused primarily by precipitation declines across the entire UCRB but higher temperatures caused about half of the 2000–2014 flow loss. The Millennium Drought was also caused by precipitation reductions in the four most productive subbasins. We evaluated the UCRB April–July runoff forecast for 2017, which decreased dramatically as the runoff season progressed. The late winter and spring 2017 was anomalously warm, but most of the reduction was due to late season dryness. The naturalized flow of the Colorado River has decreased about 15% over the last 100 yearsAbout half of the long‐term trend is attributable to rising temperatures with the remainder due to changes in precipitation patterns and other factorsSlightly over half of the flow reduction during the post‐2000 Millennium Drought is attributable to anomalously warm temperatures

In the western United States, the seasonal phase of snow storage bridges between winter‐dominant precipitation and summer‐dominant water demand. The critical role of snow in water supply has been frequently quantified using the ratio of snowmelt‐derived runoff to total runoff. However, current estimates of the fraction of annual runoff generated by snowmelt are not based on systematic analyses. Here based on hydrological model simulations and a new snowmelt tracking algorithm, we show that 53% of the total runoff in the western United States originates as snowmelt, despite only 37% of the precipitation falling as snow. In mountainous areas, snowmelt is responsible for 70% of the total runoff. By 2100, the contribution of snowmelt to runoff will decrease by one third for the western U.S. in the Intergovernmental Panel on Climate Change Representative Concentration Pathway 8.5 scenario. Snowmelt‐derived runoff currently makes up two thirds of the inflow to the region's major reservoirs. We argue that substantial impacts on water supply are likely in a warmer climate. We estimate that 53% of the total runoff in the western U.S. originates as snow, despite only 37% of the precipitation falling as snowSnowmelt produces 70% of the total runoff in mountainous areas in the WestThe ratio of snow‐derived runoff to the total runoff will reduce by one third by 2100 in a business‐as‐usual climate scenario

We estimate net groundwater storage change in the Central Valley from April 2002 to September 2016 as the difference between inflows and outflows, precipitation, evapotranspiration, and changes in soil moisture and surface water storage. We also estimate total water storage change attributable to groundwater change using Gravity Recovery and Climate Experiment (GRACE) satellite data, which should be equivalent to our water balance estimates. Over two drought periods within our 14‐1/2 years study period (January 2007 to December 2009 and October 2012 to September 2016), we estimate from our water balance that a total of 16.5 km3and 40.0 km3of groundwater was lost, respectively. Our water balance‐based estimate of the overall groundwater loss over the 14‐1/2 years is −20.7 km3, which includes substantial recovery during nondrought periods The estimated rate of groundwater loss is greater during the recent drought (10.0 ± 0.2 versus 5.5 ± 0.3 km3/yr) than in the 2007–2009 drought, due to lower net inflows, a transition from row crops to trees, and higher crop water use, notwithstanding a reduction in irrigated area. The GRACE estimates of groundwater loss (−5.0 km3/yr for both water balance and GRACE during 2007–2009, and −11.2 km3/yr for GRACE versus −10 km3/yr for water balance during 2012–2016) are quite consistent for the two methods. However, over the entire study period, the GRACE‐based groundwater loss estimate is almost triple that from the water balance, mostly because GRACE does not indicate the between‐drought groundwater recovery that is inferred from our water balance. About 55 km3of groundwater was lost from the Central Valley during the 2007–2009 and 2012–2016 droughtsReduced net inflows, transition from row to tree crops, and higher evapotranspiration lead to most of the groundwater loss in 2012–2016Water balance and GRACE estimates of CV groundwater loss agree quite closely during drought (but not during nondrought) periods

New data sources are revolutionizing the hydrological sciences. The capabilities of hydrological models have advanced greatly over the last several decades, but until recently model capabilities have outstripped the spatial resolution and accuracy of model forcings (atmospheric variables at the land surface) and the hydrologic state variables (e.g., soil moisture; snow water equivalent) that the models predict. This has begun to change, as shown in two examples here: soil moisture and drought evolution over Africa as predicted by a hydrology model forced with satellite‐derived precipitation, and observations of snow water equivalent at very high resolution over a river basin in California's Sierra Nevada. New observations are revolutionizing hydrologyDrought monitoring is now routinely performed globally using macroscale hydrological modelsFor the first time, it is now possible to observe snow water equivalent in mountainous terrain at very high‐spatial resolution

Developing interdisciplinary methods to measure ecosystem services is a scientific priority, however, progress remains slow in part because we lack ecological production functions (EPFs) to quantitatively link ecohydrological processes to human benefits. In this study, we tested a new approach, combining a process‐based model with regression models, to create EPFs to evaluate water storage and local climate regulation from a green infrastructure project on the Yongding River in Beijing, China. Seven artificial lakes and wetlands were established to improve local water storage and human comfort; evapotranspiration (ET) regulates both services. Managers want to minimize the trade‐off between water losses and cooling to sustain water supplies while lowering the heat index (HI) to improve human comfort. We selected human benefit indicators using water storage targets and Beijing's HI, and the Variable Infiltration Capacity model to determine the change in ET from the new ecosystems. We created EPFs to quantify the ecosystem services as marginal values [Δfinal ecosystem service/Δecohydrological process]: (1) Δwater loss (lake evaporation/volume)/Δdepth and (2) Δsummer HI/ΔET. We estimate the new ecosystems increased local ET by 0.7 mm/d (20.3 W/m2) on the Yongding River. However, ET rates are causing water storage shortfalls while producing no improvements in human comfort. The shallow lakes/wetlands are vulnerable to drying when inflow rates fluctuate, low depths lead to higher evaporative losses, causing water storage shortfalls with minimal cooling effects. We recommend managers make the lakes deeper to increase water storage, and plant shade trees to improve human comfort in the parks. A challenge to measuring ecosystem services is creating ecological production functions linking ecohydrological processes to human benefitsWater storage and local climate regulation were evaluated from seven artificial lakes and wetlands in BeijingLakes and wetlands increased local evapotranspiration causing water storage reductions while providing minimal cooling for human comfort

Summer streamflow variations strongly affect water supply reliability and ecological functioning of western U.S. (WUS) streams. Traditional snow‐based forecasts of summer streamflow are becoming less accurate with warming‐induced reductions in winter snow accumulation. This reflects a rising importance of competing runoff‐generating processes in controlling summer streamflow variations, primarily an increasing role of rainfall in contrast to snowmelt. Here, based on a snowmelt‐rainfall tracking algorithm applied to two hydrological models, we show that cool‐season rainfall provides an important volumetric contribution to summer streamflow for many WUS streams in the current climate, and this contribution will increase under climate warming, especially in years with warm snow droughts and abnormally dry summers. We also show that seasonal rainfall (warm‐/cool‐seasons) dominates the variability of summer streamflow across ∼70% area of WUS. We show that an increasing warm‐season rainfall contribution to summer streamflow (largely replacing snowmelt) results in reduced summer streamflow predictability. Summer streamflow is a critical water resource in the generally dry summers of the western U.S. (WUS), and is routinely forecasted using spring snowpack and/or winter total precipitation as primary predictors. However, climate warming leads to reduced snowpacks, exacerbates summer low flows, and reduces the accuracy of snow‐based summer streamflow forecasts. On the other hand, the role of winter rainfall as a control on summer streamflow increases in a warmer climate. Here, we explicitly quantify the contributions from cool‐season rainfall, warm‐season rainfall, and snowmelt to summer streamflow across the WUS, and how they change under a uniformly 1°C warmer climate. We show that the cool‐season rainfall contribution to summer streamflow increases under warming across WUS, especially in streams that currently have low‐to‐moderate snow contributions to runoff, and in years with anomalously warm winters and/or dry summers. We also show that the warm‐season rainfall contribution to summer streamflow increases widely, especially in the southern interior of WUS in a warmer climate, and that increasing warm‐season rainfall contribution to summer streamflow (largely replacing snowmelt) results in reduced summer streamflow predictability. The cool‐season rainfall contribution to summer streamflow is greatest in low‐elevation coastal streams with dry summersClimate warming leads to an increased contribution of seasonal rainfall to summer streamflow as spring snowmelt contributions declineSummer streamflow predictability declines with reduced snowmelt and increased warm‐season rainfall contribution in a warmer climate The cool‐season rainfall contribution to summer streamflow is greatest in low‐elevation coastal streams with dry summers Climate warming leads to an increased contribution of seasonal rainfall to summer streamflow as spring snowmelt contributions decline Summer streamflow predictability declines with reduced snowmelt and increased warm‐season rainfall contribution in a warmer climate

In many partially glacierized watersheds glacier recession driven by a warming climate could lead to complex patterns of streamflow response over time, often marked with rapid increases followed by sharp declines, depending on initial glacier ice cover and rate of climate change. Capturing such “phases” of hydrologic response is critical in regions where communities rely on glacier meltwater, particularly during low flows. In this paper, we investigate glacio‐hydrologic response in the headwaters of the Zongo River, Bolivia, under climate change using a distributed glacio‐hydrological model over the period of 1987–2100. Model predictions are evaluated through comparisons with satellite‐derived glacier extent estimates, glacier surface velocity, in situ glacier mass balance, surface energy flux, and stream discharge measurements. Historically (1987–2010) modeled glacier melt accounts for 27% of annual runoff, and 61% of dry season (JJA) runoff on average. During this period the relative glacier cover was observed to decline from 35 to 21% of the watershed. In the future, annual and dry season discharge is projected to decrease by 4% and 27% by midcentury and 25% and 57% by the end of the century, respectively, following the loss of 81% of the ice in the watershed. Modeled runoff patterns evolve through the interplay of positive and negative trends in glacier melt and increased evapotranspiration as the climate warms. Sensitivity analyses demonstrate that the selection of model surface energy balance parameters greatly influences the trajectory of hydrological change projected during the first half of the 21st century. These model results underscore the importance of coupled glacio‐hydrology modeling. Representation ice flow is critical to depict phased hydrologic response to glacier recessionPredictions of glacier and runoff patterns are highly sensitive to the selection of energy balance parametersIncreased evapotranspiration is critical to consider in long‐term projections of water availability

Flash drought is a term that was popularized during rapidly evolving droughts in the Central U.S. in 2012 that were associated with heat waves. We posit that there are two kinds of flash droughts, and we will focus on heat wave flash droughts, of which the 2012 events were typical. We find, based on an analysis of temperature observations and model‐reconstructed soil moisture (SM) and evapotranspiration from 1916 to 2013, that heat wave flash droughts in the conterminous U.S. (CONUS) are most likely to occur over the Midwest and the Pacific Northwest during the growing season. We also find that the number of such events across the CONUS has been decreasing over the last century but rebounded after 2011. The long‐term downward trends appear to be associated with generally increasing trends in SM resulting from increasing trends in precipitation over the areas where heat wave flash droughts are most likely to occur. Definition of flash droughtFlash droughts are in declineDroughts occur less often and cover smaller areas

California has experienced severe drought in 2012–2014 (which appears to be continuing into 2015), with especially low winter precipitation and mountain snowpack in winter 2013–2014. However, the extent to which climate change is implicated in the drought, if at all, is not clear. By applying modeling and statistical approaches, we construct a historical record of California snowpack, runoff, and other hydrological variables of almost 100 years in length and use the reconstructed records to analyze climate trends in the Sierra Nevada and their impact on extreme drought events in the historic record. We confirm a general warming trend and associated decreasing trends in spring snowpack and runoff. We find that the warming may have slightly exacerbated some extreme events (including the 2013–2014 drought and the 1976–1977 drought of record), but the effect is modest; instead, these drought events are mainly the result of variability in precipitation. 2013–2014 was a dry year, but far less so than 1976–1977The 2 year dry period 2012–2014 likewise was anomalous but far less so than 1975–1977The variability in precipitation is the main cause of the drought

Increased temperatures and changes in precipitation will result in fundamental changes in the seasonal distribution of streamflow in the Pacific Northwest and will have serious implications for water resources management. To better understand local impacts of regional climate change, we conducted model experiments to determine hydrologic sensitivities of annual, seasonal, and monthly runoff to imposed annual and seasonal changes in precipitation and temperature. We used the Variable Infiltration Capacity (VIC) land‐surface hydrology model applied at 1/16° latitude‐longitude spatial resolution over the Pacific Northwest (PNW), a scale sufficient to support analyses at the hydrologic unit code eight (HUC‐8) basin level. These experiments resolve the spatial character of the sensitivity of future water supply to precipitation and temperature changes by identifying the seasons and locations where climate change will have the biggest impact on runoff. The PNW exhibited a diversity of responses, where transitional (intermediate elevation) watersheds experience the greatest seasonal shifts in runoff in response to cool season warming. We also developed a methodology that uses these hydrologic sensitivities as basin‐specific transfer functions to estimate future changes in long‐term mean monthly hydrographs directly from climate model output of precipitation and temperature. When principles of linearity and superposition apply, these transfer functions can provide feasible first‐order estimates of the likely nature of future seasonal streamflow change without performing downscaling and detailed model simulations. Hydrologic sensitivities characterize seasons/locations susceptible to future changeTransitional watersheds experience greatest seasonal shifts in runoffSeasonal streamflow estimation works well when linearity and superposition apply

A recent report of the Interagency Advisory Committee on Water Data found that there is “no current procedure for assigning an exceedance probability to the probable maximum flood PMF… in a reliable, consistent, or credible manner.” This conclusion was used as justification for continuation of the current, quasideterministic, PMFbased spillway design methods used by all federal agencies. This is despite criticism by both researchers and practitioners that PMFbased methods tend to lead to a false sense of security and to misallocation of resources for dam safety improvements. As an alternative to perpetuation of the status quo, the writers outline four general areas in which research should be promoted for improved estimation of extreme floods, as well as research aimed at development of a method for incorporation of risk information into spillway design. The writers believe that if the federal action agencies were to promote research in the areas suggested, the current stagnation that has set in would be broken, and the selffulfilling prophecy that there are no alternatives to current practice could no longer be justified.

Simulation models have proven a durable and useful tool for the development of water resource system operating policies. One important component of any operating policy is streamflow forecasting. Incorporation of streamflow forecasts in a simulation model can be difficult because of the stochastic nature of forecasts and the range of forecast models in use. In this work, a forecast generation algorithm is developed which is applicable for any forecast model, and requires as input only the coefficient of prediction, a measure of forecast accuracy, and the first three moments of the forecast period flows. The algorithm uses a lag one Markov structure for forecast errors, where the correlation coefficient is a function of the forecast model coefficient of prediction. The generation algorithm is demonstrated for simulation of 7 and 120 day forecasts for the Skykomish River, Washington.

We estimated runoff sensitivities to global mean temperature change using climate experiments archived in the Coupled Model Intercomparison Project Phase 5 (CMIP5) and compared with the similar CMIP3 results. We also evaluated differences in runoff sensitivity for the Earth System Models (ESMs) and Climate System Models (CSMs), which were separately identified in CMIP5 for the first time. Our results show that runoff sensitivity is relatively independent of emission scenarios in CMIP5, as in CMIP3 results. Global mean runoff would increase about 2.9% per °C of global warming in CMIP5, as contrasted with 1.9% in CMIP3. Although global runoff sensitivity patterns for CMIP5 and CMIP3 are roughly similar, CMIP5 suggests significant declines (increases) in runoff sensitivity over 17% (25%) of the global land area relative to CMIP3. Globally, the ESMs and CSMs have about the same model spread in runoff change projections. CMIP5 results imply global runoff increases by 2.9% per degree C global warmingEarth System Models do not reduce model spread in runoff change projectionsRunoff sensitivities in CMIP5 models are consistent across RCP scenarios

In the United States and Europe, at present 91% and 78% (ref. ) of the total electricity is produced by thermoelectric (nuclear and fossil-fuelled) power plants, which directly depend on the availability and temperature of water resources for cooling. During recent warm, dry summers several thermoelectric power plants in Europe and the southeastern United States were forced to reduce production owing to cooling-water scarcity. Here we show that thermoelectric power in Europe and the United States is vulnerable to climate change owing to the combined impacts of lower summer river flows and higher river water temperatures. Using a physically based hydrological and water temperature modelling framework in combination with an electricity production model, we show a summer average decrease in capacity of power plants of 6.3–19% in Europe and 4.4–16% in the United States depending on cooling system type and climate scenario for 2031–2060. In addition, probabilities of extreme (>90%) reductions in thermoelectric power production will on average increase by a factor of three. Considering the increase in future electricity demand, there is a strong need for improved climate adaptation strategies in the thermoelectric power sector to assure futureenergy security.

Climate change is linked to changing precipitation, yet uncertainty remains as to whether and how hydroclimatic changes will influence flood frequency, magnitude, and timing. Part of this uncertainty is because few regional‐scale studies have analyzed changes in floods produced by different mechanisms. We identify six flood generating mechanisms (FGMs) based on prominent meteorological processes and moisture supply to the soil column in the Western U.S.; these include large‐scale frontal storms (atmospheric and non‐atmospheric rivers), monsoons, convective storms, snowmelt, and rain‐on‐snow. We assess trends in the frequency, magnitude, and timing of floods produced by different FGMs across 119 basins during 1960–2018. Overall, we find decreased frequency and magnitude of rain‐on‐snow‐driven floods, increased frequency and magnitude of convective‐storm‐driven floods, and a temporal shift to earlier in the year for snowmelt‐driven floods. The flood characteristics are generally stable for floods produced by other FGMs. Floods are among the most common and costliest of natural disasters. Extreme rainfall is a major cause of flooding and has become more common as global temperatures have increased. However, many studies show that the number and size of flood events remain mostly unchanged. It is possible that climate change is affecting different types of floods in different ways, and so we separate floods into different categories and evaluate trends on a category‐by‐category basis. In particular, we group floods across the western U.S. into six categories, which we term flood generating mechanisms (FGMs). These include rain‐on‐snow, snowmelt, monsoons, convective storms, and two types of frontal storms. We then separately describe trends in the frequency, magnitude, and timing for each FGM from 1960 to 2018. We find that floods generated by convective storms have become more common and more extreme. On the other hand, rain‐on‐snow floods have become rarer and less extreme. Also, floods caused by snowmelt are occurring earlier within the year. Most of the other categories do not show significant changes in flood characteristics. These results give us important information on the types of flooding that may be most impacted by future climate change, which can aid future flood planning efforts. The frequency and magnitude of rain‐on‐snow floods have decreased across the western U.S., and snowmelt floods are occurring earlierFloods generated by convective storms (primarily in the Southwest) have increased in frequency and magnitudeExcept for the above‐mentioned changes, the frequency, magnitude, and timing of floods have been relatively stable over time The frequency and magnitude of rain‐on‐snow floods have decreased across the western U.S., and snowmelt floods are occurring earlier Floods generated by convective storms (primarily in the Southwest) have increased in frequency and magnitude Except for the above‐mentioned changes, the frequency, magnitude, and timing of floods have been relatively stable over time

Future climate warming over the Western U.S. (WUS) is projected to be greater in summer than winter. Previous model‐based studies of large river basins in the WUS showed much different annual streamflow responses to warming in warm versus cool seasons. However, it remains unclear how the relative responses annual streamflow relative responses to seasonal warming (asymmetry) and drivers of the response asymmetry vary across the entire WUS at the catchment‐scale, and how the simulated results compare with observations. Here, we investigate the asymmetry of annual streamflow responses to warm versus cool season warming at the HUC‐8 level across the entire WUS using model simulations and observations. We also examine the asymmetries' relationship with land surface and hydroclimate characteristics, and the primary contributor to the response asymmetry for each HUC‐8 basin. The HUC‐8 level results reveal more complexity than do earlier analyses of much larger river basins. Over 25% of WUS area has annual streamflow increases in response to warming in at least one season (mostly cool season). Annual streamflow is most sensitive to warm season warming in cool, inland basins, especially the northern Columbia River basin and most of the Upper Colorado River Basin, and most sensitive to cool season warming in warm, coastal basins. This bi‐directional pattern is enhanced by vegetation coverage but weakened by long‐term snowpack decline. In coldest basins with short snow‐free seasons, net radiation changes dominate the streamflow response asymmetry. For basins with cold‐to‐intermediate temperatures, vapor pressure deficit changes dominate. For warmest basins, surface resistance changes dominate. Annual streamflow relative response to seasonal warming (asymmetry) is similar between large and HUC‐8 basins across the Western U.SVegetation enhances, while long‐term snowpack‐decline reduces, the asymmetry of annual streamflow response to seasonal warmingVapor pressure deficit and surface resistance dominate asymmetry aside from a few of the coldest basins where radiation controls Annual streamflow relative response to seasonal warming (asymmetry) is similar between large and HUC‐8 basins across the Western U.S Vegetation enhances, while long‐term snowpack‐decline reduces, the asymmetry of annual streamflow response to seasonal warming Vapor pressure deficit and surface resistance dominate asymmetry aside from a few of the coldest basins where radiation controls

Chinook salmon (Oncorhynchus tshawytscha), hatchery-reared as juveniles, returned to the upper Columbia River Basin in numbers exceeding broodstock and fishery needs during the spring of 2000. Plans to euthanize these adults were opposed by some regional stakeholders, who preferred letting them spawn naturally in streams also used by endangered spring-run chinook salmon. The National Marine Fisheries Service requested that the Independent Scientific Advisory Board review the scientific literature and conclude whether it was biologically sound to permit hatchery-origin adult salmon to spawn in the wild in large numbers. Substantial experimental evidence demonstrates that domestication selection can genetically alter hatchery populations in a few generations and that hatchery-origin adults returning from the ocean and spawning in the wild produce fewer progeny than adults of wild origin spawning in the wild. More limited evidence suggests that interbreeding between hatchery-origin adults and wild fish can reduce the fitness of the wild population. We conclude that decisions whether or not to permit hatchery-origin adults to spawn in the wild should be based on the needs of wild populations and the ability of the habitat to support additional reproduction, not based simply on the availability of hatchery-origin adults returning from the ocean.

Focuses on the use of advances in climate research to improve streamflow forecasts at seasonal-to-interannual, decadal, and longer time scales. Improvement in the understanding of, and an ability to forecast, El Nino/Southern Oscillation at seasonal/interannual time scales; Development of a streamflow forecasting techniques for Pacific Northwest rivers; Use of the data meteorological from previous years.

Groundwater depletion is a major threat to agricultural and municipal water supply in California's Central Valley. Recent droughts during 2007–2009 and 2012–2016 exacerbated chronic groundwater depletion. However, it is unclear how much groundwater storage recovered from drought‐related overdrafts during post‐drought years, and how climatic conditions and water management affected recovery times. We estimated groundwater storage change in the Central Valley for April 2002 through September 2019 using four methods: GRACE satellite data, a water balance approach, a hydrologic simulation model, and monitoring wells. We also evaluated the sensitivity of drought recovery to different climate scenarios (recent climate ± droughts and future climate change scenarios: 20 GCMs and 2 RCPs) using water balance method and statistical sampling of historical climate data. Estimated Central Valley groundwater loss during the two droughts ranged from 19 km3(2007–2009) to 28 km3(2012–2016) (median of four methods). Median aquifer storage recovery was 34% and 19% of the overdraft during the 2010–2011 and 2017–2019 post‐drought years, respectively. Numerical experiments show that recovery times are sensitive to climate forcing, with longer recovery times for a future climate scenario that replicate historical climatology relative to historical forcing with no droughts. Overdraft recovery times decrease by ∼2× with implementation of pumping restrictions (30th to 50th percentiles of historical groundwater depletion) to constrain groundwater depletion relative to no restrictions with a no‐drought future climatology. This study highlights the importance of considering water management implications for future drought recoveries within the context of climate change scenarios. California's Central Valley has experienced chronic groundwater depletion over the past few decades, the rate of which was amplified by droughts in 2007–2009 and 2012–2016. There is limited knowledge as to how much of the drought‐caused groundwater depletion has recovered during post‐drought years and how climate and water management affect overdraft recovery times. We address these issues by estimating groundwater storage changes using four methods and conducting numerical experiments with varying climatic conditions and water management options. We find that less than one‐third of the groundwater overdraft from the most recent droughts was recovered during post‐drought years. Projected overdraft recovery times vary greatly depending on the climate scenarios and water management strategies, and future droughts are likely to cause overdrafts from which recovery is unlikely given the current level of groundwater extractions. However, management measures such as capping groundwater pumping could reduce recovery times by a factor of two or more depending on the groundwater extraction cap and post‐drought climate. Groundwater storage recovery during post‐drought periods ranged from 34% (2007–2009 drought) to 19% (2012–2016 drought)Projected drought recovery times decrease by a factor of 3.6–7.8 with post‐drought periods containing no drought years or wet years onlyOverdraft recovery time decreased by a factor of 2 with implementation of modest pumping restrictions under no‐drought post‐drought climate Groundwater storage recovery during post‐drought periods ranged from 34% (2007–2009 drought) to 19% (2012–2016 drought) Projected drought recovery times decrease by a factor of 3.6–7.8 with post‐drought periods containing no drought years or wet years only Overdraft recovery time decreased by a factor of 2 with implementation of modest pumping restrictions under no‐drought post‐drought climate

We examine the characteristics of extremely severe and extensive floods across the Western U.S., which we term superfloods. We develop a system to score each day from 1950 to 2010, across the conterminous U.S. west of the Continental Divide, according to the severity of flooding and the extent of the area affected on that day. In order to augment the available stream gauge data, we incorporate model output runoff from the Variable Infiltration Capacity model aggregated to the HUC‐8 basin level. The superfloods we identify include well‐known events such as the Christmas floods of December 1964. Once the superfloods are identified, we use modeled snow water equivalent and snowmelt in combination with gridded observed precipitation and air temperature to identify superflood drivers. We find that the superfloods fall into three distinct categories: (a) winter events, most of which are associated with Atmospheric Rivers, many enhanced by rain‐on‐snow; (b) late spring floods associated with snowmelt, including some events with rainfall contributions; and (c) rainfall‐driven warm season floods associated with tropical storms (primarily in the Southwest). We find that rainfall is often the primary driver of superfloods, although snowmelt also contributes in rain‐snow transition zones. Regardless of superflood characteristics, extreme conditions in combination are a common feature, including combinations of rainfall and snowmelt as well as successive rainstorms. Our results provide a regional view of flood drivers and how they have combined to produce superfloods across the region over our 61‐year study period. We identify superfloods in the Western U.S. as historical floods of exceptional magnitudes that covered large areasSuperfloods categories are: winter (mostly Atmospheric Rivers), spring snowmelt (some with rainfall) and summer (tropical storms)Superfloods are caused both by extreme precipitation or snowmelt and unusual combinations of other controlling factors We identify superfloods in the Western U.S. as historical floods of exceptional magnitudes that covered large areas Superfloods categories are: winter (mostly Atmospheric Rivers), spring snowmelt (some with rainfall) and summer (tropical storms) Superfloods are caused both by extreme precipitation or snowmelt and unusual combinations of other controlling factors

We utilized a macroscale hydrology model in conjunction with an atmospheric river (AR) catalog to evaluate AR‐affected snow accumulation in the Upper Colorado River basin (UCRB) headwaters for water years 1951–2015. We find that there are on average 11.4 days during which AR‐originating snow accumulates in the UCRB each year, yielding 5.75 km3of snow water equivalent (SWE). This AR‐originating SWE accounts for 30.8% of average annual peak SWE in the basin. Most of the ARs that reach the UCRB first intercept the Sierra Nevada region of California; 84% of days affected by ARs in the UCRB are also AR‐affected days in the Sierra Nevada. The pathways of the ARs reaching the UCRB do not have a great effect on snow accumulation, however ARs that first pass through the Sierra Nevada generally produce greater snow accumulations there than in the UCRB. Atmospheric rivers (AR), commonly defined as corridors of concentrated water vapor in the atmosphere, can yield large amounts of snow accumulation when they make landfall during the cold season. As over half of the naturalized streamflow in the Colorado River originates from water released by snow melt, it is of great significance to understand how atmospheric rivers can affect the snowpack in the basin. Here, we identify the atmospheric rivers in the Upper Colorado River basin during a 65‐years‐long historical record and evaluate their contribution to mountain snowpack. Almost one third of snowpack in the basin is attributed to atmospheric river induced snowfall. The primary origin of the relevant ARs is from the southwest, although the ARs' pathways do not much affect the amount of snow they yield. Non‐AR‐related snowfall is highly important to the basin's snow accumulation during cold years, while AR contributions to snow are greater under warm conditions. Atmospheric river (AR) induced snow accumulation accounts for 30.8% of the peak snow water equivalent in the Upper Colorado River basin (UCRB)Most of the ARs that reach the UCRB first intercept the Sierra Nevada RegionAside from very dry conditions, the AR contribution to UCRB snowpack during warm winters is greater than average Atmospheric river (AR) induced snow accumulation accounts for 30.8% of the peak snow water equivalent in the Upper Colorado River basin (UCRB) Most of the ARs that reach the UCRB first intercept the Sierra Nevada Region Aside from very dry conditions, the AR contribution to UCRB snowpack during warm winters is greater than average

AbstractSpatial patterns in trends of four monthly variables: average temperature, precipitation, streamflow, and average of the daily temperature range were examined for the continental United States for the period 1948–88. The data used are a subset of the Historical Climatology Network (1036 stations) and a stream gage network of 1009 stations. Trend significance was determined using the nonparametric seasonal Kendall's test on a monthly and annual basis, and a robust slope estimator was used for determination of trend magnitudes. A bivariate test was used for evaluation of relative changes in the variables, specifically, streamflow relative to precipitation, streamflow relative to temperature, and precipitation relative to temperature.Strong trends were found in all of the variables at many more stations than would be expected due to chance. There is a strong spatial and seasonal structure in the trend results. For instance, although annual temperature increases were found at many stations, mostly in the North and West, there were almost as many downtrends, especially in the South and East. Among the most important trend patterns are (a) increases in March temperature at almost half of the stations; (b) increases in precipitation from September through December at as many as 25 percent of the stations, mostly in the central part of the country; (c) strong increases in streamflow in the period November–April at a maximum of almost half of the stations, with the largest trend magnitudes in the north-central states; (d) changes in the temperature range (mostly downward) at a large number of stations beginning in late spring and continuing through winter, affecting as many as over half of the stations. The observed trends in streamflow are not entirely consistent with the changes in the climatic variables and may be due to a combination of climatic and water management effects.

AbstractThe sensitivity of global climate to the characterization of the land-surface hydrology is investigated using the Geophysical Fluid Dynamics Laboratory GCM at R15 resolution with the standard Budyko bucket and the Variable Infiltration Capacity (VIC) Model, which incorporates a parsimonious parameterization of the subgrid-scale spatial variability of soil moisture capacity, as well as base flow, by means of soil moisture drainage during dry periods. Four experiments were performed using the VIC model. The first used a globally fixed soil moisture capacity of 15 cm to provide a comparison to the Budyko bucket. The second used a more realistic globally varying soil moisture capacity. The third and fourth were sensitivity experiments using globally fixed soil moisture capacities of 5 and 25 cm. The results of the VIC fixed runs (15 cm) showed that global average soil moisture was considerably lower (about 2.5 cm on average) as compared with the bucket runs, global evaporation and precipitation were reduced, and surface air temperature was increased, especially in the Northern Hemisphere in summer. The greater sensitivity of the Northern Hemisphere land areas to the altered land hydrology is attributed primarily to recycling of summertime precipitation in the interior of these continents. The authors found, somewhat surprisingly, that the water-holding capacities of the VIC model had relatively little influence .on the simulated climates of northern Eurasia and North America. This is attributed to the fact that much of the soil moisture capacity is unutilized for evaporation, due to the dry period drainage to base flow. The results argue for representation of the surface hydrology in GCMs with two-layer sail models, which are capable of representing the cycling of moisture during dry periods by means of surface evaporation, which is generally underestimated by single-layer models.

Intervention analysis is a rigorous statistical method for analyzing the effects of man-induced or natural changes on the environment. For instance, it may be necessary to determine whether a newly installed pollution control device significantly reduces the former mean level of a pollutant. By using intervention analysis, the actual change in the pollutant levels can be statistically determined. Previously, no comprehensive method was available to assess changes in the environment. Intervention analysis is an advanced type of Box-Jenkins model. A genpral description of Box-Jenkins models and their extensions is given. Also, the importance of adhering to sound modeling principles when fitting a stochastic model to a time series is emphasized. Following a discussion of intervention models, three applications of intervention analysis to environmental problems are given. Two applications deal with the environmental effects of man-made projects, while the third example demonstrates how a forest fire can affect the flow regime of a river.

Intervention analysis is a relatively new branch of time series analysis. The power of this technique, which gives the probability that changes in mean level can be distinguished from natural data variability, is quite sensitive to the way the data are collected. The principal independent variables influenced by the data collection design are overall sample size, sampling frequency, and the relative length of record before the occurrence of the event (intervention) that is postulated to have caused a change in mean process level.

The possible effects of climate change on the combined Central Valley Project‐California State Water Project (CVP/SWP) were evaluated using a three‐stage approach. In the first stage, runoff from four headwater “study catchments” was simulated using rainfall/snowmelt‐runoff models, with climatic input taken from CO2 doubling scenarios from three general circulation models (GCMs). In the second stage, long‐term inflows to the CVP/SWP reservoir system were simulated, conditioned on the study catchment flows, using a stochastic disaggregation model. In the third stage, a system simulation model was used to evaluate the performance of the reservoir system. For all of the alternative climate scenarios, runoff would be shifted from the spring to the winter. Significantly lower water deliveries from the SWP would occur under the CO2 doubling scenarios. The reduced deliveries would occur because some of the increased winter runoff would be spilled from the reservoirs instead of being stored in the snowpack, even though the mean annual runoff increased slightly under some climate scenarios. Annual San Francisco Bay delta flows would increase under all three climate scenarios; however, flows to the bay would be substantially increased in winter and somewhat decreased in spring and summer.

Forecasting of seasonal snowmelt runoff is an important exception to the otherwise limited accuracy of long‐term runoff forecasts. Conceptual models of the snow accumulation and ablation and precipitation‐runoff processes have been advocated as a means of introducing physical understanding of the dominant hydrologic processes into seasonal snowmelt runoff processes. However, the paucity of information and extreme spatial variability of meteorological processes and watershed characteristics in mountainous watersheds often result in large simulation errors. The analysis reported here relates monthly simulation error statistics to maximum achievable seasonal runoff forecast accuracy, presuming perfect knowledge of future (gage) precipitation and temperature, but without updating of model parameters or states. In some cases, this accuracy limit is less than that actually achievable using simpler models. The results suggest a preference for simple models if the coefficient of variation of monthly simulation error cannot be reduced below about 10–15&percent;.

A weather classification scheme was coupled with a semi-Markov model to represent the coincident occurrence of rain/no rain states at a single rain gauge and classes representing regional atmospheric circulation patterns, as identified from National Meteorological Center gridded observations for a large area of the North Pacific. Weather classes were identified from daily observations of surface pressure and 850 mb pressure height at five selected ten degree latitude by ten degree longitude cells using a K-means clustering algorithm, which was applied on a month-by-month basis. The number of climate classes, K, for each month was chosen based on a preliminary analysis of the model's ability to describe statistics of observed precipitation occurrences at the Stampede Pass, Washington weather station. The length of stay distributions within each precipitation occurrence/weather class were assumed to be geometric, and the precipitation amounts for each class and season were fitted with a mixed exponential distribution. Parameters of the length of stay distributions, transition probabilities, and precipitation amounts were estimated from the period of record 1975–84.

The Seattle Water Department receives its supply from reservoirs on the Cedar and Tolt rivers, which drain the western slopes of the Cascade Mountains. Most droughts in the Northwest are the result of deficient winter snowfall. The 1987 drought was unusual in that the previous winter's precipitation was close to normal, but the summer dry period extended into the late fall, causing a severe water shortage. For this reason, a medium‐term forecast model was developed to assist in the evaluation of management alternatives during a critical 16‐week period from late summer to late fall 1987. A key feature of the model was that it predicted a best forecast at each of four inflow points, as well as the probability distribution of the forecasts.

The Boreal Ecosystem Atmosphere Study (BOREAS) is a largescale international field experiment that has the goal of improving our understanding of the exchanges of radiative energy, heat, water, CO2, and trace gases between the boreal forest and the lower atmosphere. An important objective of BOREAS is to collect the data needed to improve computer simulation models of the processes controlling these exchanges so that scientists can anticipate the effects of global change.From August 1993 through September 1994, a continuous set of monitoring measurements—meteorology, hydrology, and satellite remote sensing—were gathered overthe 1000 × 1000 km BOREAS study region that covers most of Saskatchewan and Manitoba, Canada. This monitoring program was punctuated by six campaigns that saw the deployment of some 300 scientists and aircrew into the field, supported by 11 research aircraft. The participants were drawn primarily from U.S. and Canadian agencies and universities, although there were also important contributions from France, the United Kingdom, and Russia. The field campaigns lasted for a total of 123 days and saw the compilation of a comprehensive surfaceatmosphere flux dataset supported by ecological, trace gas, hydrological, and remote sensing science observations. The surface-atmosphere fluxes of sensible heat, latent heat, CO2, and momentum were measured using eddy correlation equipment mounted on a surface network of 10 towers complemented by four flux-measurement aircraft. All in all, over 350 airborne missions (remote sensing and eddy correlation) were flown during the 1994 field year.Preliminary analyses of the data indicate that the area-averaged photosynthetic capacity of the boreal forest is much less than that of the temperate forests to the south. This is reflected in very low photosynthetic and carbon drawdown rates, which in turn are associated with low transpiration rates (less than 2 mm day−1 over the growing season for the coniferous species in the area). The strong sensible fluxes generated as a result of this often lead to the development of a deep dry planetary boundary layer overthe forest, particularly during the spring and early summer. The effects of frozen soils and the strong physiological control of evapotranspiration in the biome do not seem to be well represented in most operational general circulation models of the atmosphere.Analyses of the data will continue through 1995 and 1996. Some limited revisits to the field are anticipated.

AbstractThe results of simulation experiments are reported in which two 1° lat × 1° long regions were discretized into pixels of size roughly 180 m × 120 m and modeled using a hydrologically based, spatially distributed water and energy balance model. Fluxes aggregated from the distributed model (ADM), and computed using a macroscale equivalent model (MSE), which treats the entire region as a point, were compared for 2 years for two regions in Montana: one in the mountainous, semihumid western part of the state, and another in the drier, less mountainous east. The forcings for MSE were the spatial averages of precipitation, downward shortwave and longwave radiation, air temperature, wind, and vapor pressure over the respective regions spatially averaged from the distributed model.In the western region, major differences in predicted snow water equivalent between ADM and MSE were observed during the spring snowmelt period, primarily due to snow at high elevations, which is not represented by MSE. These differences persisted for smaller 0.2° × 0.2° subregions; however, an alternate probability-based partitioning of the region into 10 elevation bands greatly reduced the differences. In the eastern region, where snow accumulations are episodic, differences in snow water equivalent were due primarily to the failure of MSE to represent topographic variations in solar radiation. Differences in latent and sensible heat fluxes between ADM and MSE were greatest when MSE predicted no snow cover and ADM predicted partial area snow coverage. When both models predicted at least partial snow cover, or both models predicted no snow cover, the diurnal patterns in latent and sensible heat fluxes were similar, although MSE tended to predict larger diurnal extremes. This is attributable to the representation of partial area coverage of snow in winter and spring, and partial areas coverage of convective rainfall in summer and fall, in ADM.

A recent report of the Interagency Advisory Committee on Water Data found that there is “no current procedure for assigning an exceedance probability to the probable maximum flood (PMF)… in a reliable, consistent, or credible manner.” This conclusion was used as justification for continuation of the current, quasi‐deterministic, PMF‐based spillway design methods used by all federal agencies. This is despite criticism by both researchers and practitioners that PMF‐based methods tend to lead to a false sense of security and to misallocation of resources for dam safety improvements. As an alternative to perpetuation of the status quo, the writers outline four general areas in which research should be promoted for improved estimation of extreme floods, as well as research aimed at development of a method for incorporation of risk information into spillway design. The writers believe that if the federal action agencies were to promote research in the areas suggested, the current stagnation that has set in would be broken, and the self‐fulfilling prophecy that there are no alternatives to current practice could no longer be justified.

Accurate estimation of water supply reliability is often hindered by a paucity of historic streamflow data and the expense of adequate system modelling. A procedure using synthetic streamflow generation and screening models in determining water supply reliability in a cost-effective manner is presented. A series of screening models are developed to evaluate numerous synthetic streamflow data to select from them desired flows for incorporation into an optimization model. Results from the optimization model are used to generate cumulative distribution functions of system reliability. The water supply system serving Seattle, Washington is used as a case study.