Search

Flash droughts are a recently recognized type of extreme event distinguished by sudden onset and rapid intensification of drought conditions with severe impacts. They unfold on subseasonal-to-seasonal timescales (weeks to months), presenting a new challenge for the surge of interest in improving subseasonal-to-seasonal prediction. Here we discuss existing prediction capability for flash droughts and what is needed to establish their predictability. We place them in the context of synoptic to centennial phenomena, consider how they could be incorporated into early warning systems and risk management, and propose two definitions. The growing awareness that flash droughts involve particular processes and severe impacts, and probably a climate change dimension, makes them a compelling frontier for research, monitoring and prediction. Flash droughts, which develop over the course of weeks, are difficult to forecast given the current state of subseasonal-to-seasonal prediction. This Perspective offers operational and research definitions, places them in the broader context of climate and suggests avenues for future research.

‘Flash drought’ (FD) describes the rapid onset of drought on sub-seasonal times scales. It is of particular interest for agriculture as it can deplete soil moisture for crop growth in just a few weeks. To better understand the processes causing FD, we evaluate the importance of evaporative demand and precipitation by comparing three different drought indices that estimate this hazard using meteorological and hydrological parameters from the CMIP5 suite of models. We apply the Standardized Precipitation Index (SPI); the Evaporative Demand Drought Index (EDDI), derived from evaporative demand (E0); and the Evaporative Stress Index (ESI), which connects atmospheric and soil moisture conditions by measuring the ratio of actual and potential evaporation. The results show moderate-to-strong relationships (r2 > 0.5) between drought indices and upper level soil moisture on daily time scales, especially in drought-prone regions. We find that all indices are able to identify FD in the top 10-cm layer of soil moisture in a similar proportion to that in the models’ climatologies. However, there is significant inter-model spread in the characteristics of the FDs identified. This spread is mainly caused by an overestimation of E0, indicating stark differences in the land surface models and coupling in individual CMIP5 models. Of all indices, the SPI provides the highest skill in predicting FD prior to or at the time of onset in soil moisture, while both EDDI and ESI show significantly lower skill. The results highlight that the lack of precipitation is the main contributor to FDs in climate models, with E0 playing a secondary role.

Until the scientific community coalesces around a consensus definition of flash drought, we might usefully distinguish them from “ordinary” droughts by applying a criterion of a rapid intensification from near-normal soil moisture to drought conditions over a period of a few weeks. Here, we use such a definition to generate the first spatially distributed, long-term climatology of flash droughts across Australia, which we derive using a suite of indices that capture both the supply and the demand perspectives of drought: evaporative demand describes the atmospheric demand for moisture from the surface; precipitation, the supply of moisture from the atmosphere to the surface; and evaporative stress, the supply of moisture from the surface relative to evaporative demand.Regardless of metric-based definition, flash droughts are observed across all seasons. They can terminate as rapidly as they start, but in some cases can eventuate in a seasonal-scale drought. We show that flash-drought variability and its prevalence can be related to ENSO phases, which suggests an opportunity for enhanced seasonal-scale prediction. We examine a case study in the Wimmera Region of southeast Australia (around the South Australia / Victoria border), we show that monitoring precipitation is less useful for capturing the onset of flash drought. Instead, indices that capture the demand perspective of drought--such as the Evaporative Demand Drought Index (EDDI) and Evaporative Stress Index (ESI)--are more useful for monitoring flash-drought development.

In eastern Africa droughts can cause crop failure and lead to food insecurity. With increasing temperatures, there is an a priori assumption that droughts are becoming more severe. However, the link between droughts and climate change is not sufficiently understood. Here we investigate trends in long-term agricultural drought and the influence of increasing temperatures and precipitation deficits. Using a combination of models and observational datasets, we studied trends, spanning the period from 1900 (to approximate pre-industrial conditions) to 2018, for six regions in eastern Africa in four drought-related annually averaged variables: soil moisture, precipitation, temperature, and evaporative demand (E0). In standardized soil moisture data, we found no discernible trends. The strongest influence on soil moisture variability was from precipitation, especially in the drier or water-limited study regions; temperature and E0 did not demonstrate strong relations to soil moisture. However, the error margins on precipitation trend estimates are large and no clear trend is evident, whereas significant positive trends were observed in local temperatures. The trends in E0 are predominantly positive, but we do not find strong relations between E0 and soil moisture trends. Nevertheless, the E0 trend results can still be of interest for irrigation purposes because it is E0 that determines the maximum evaporation rate. We conclude that until now the impact of increasing local temperatures on agricultural drought in eastern Africa is limited and we recommend that any soil moisture analysis be supplemented by an analysis of precipitation deficit., Earth System Dynamics, 12 (1), ISSN:2190-4987, ISSN:2190-4979

Loss of sunlight-reflecting snow spurs evaporation and ebbs river flow

Drought is a complex phenomenon manifested through interactions between biophysical and social factors. At the Wind River Indian Reservation (WRIR) in west-central Wyoming, water shortages have become increasingly common since the turn of the 21st century. Here we discuss the 2015 water year as an exemplar year, which was characterized by wetter-than-normal conditions across the reservation and, according to the U.S. Drought Monitor, remained drought-free throughout the year. Yet parts of the reservation experienced harmful water shortages, or “micro-drought” conditions, during the growing season in 2015. In this assessment of the 2015 water year at the WRIR we: (1) describe the hydroclimatic and social processes under way that contributed to the 2015 water year micro-drought in the Little Wind Basin; (2) compare water availability conditions within and between other basins at the WRIR to illustrate how micro-droughts can result from social and environmental features unique to local systems; and (3) describe how a collaborative project is supporting drought preparedness at the WRIR. We combine a social science assessment with an analysis of the hydroclimate to deconstruct how shortages manifest at the WRIR. We provide insights from this study to help guide drought assessments at local scales. Keywords: Drought, Climate vulnerability, Drought preparedness, Indigenous adaptation, Co-production

The Famine Early Warning Systems Network (FEWS NET) team provides food insecurity outlooks for several developing countries in Africa, central Asia, and Central America. This study describes development of a new global reference evapotranspiration (ET0) seasonal reforecast and skill evaluation with a particular emphasis on the potential use of this dataset by FEWS NET to support food insecurity early warning. The ET0 reforecasts span the 1982–2009 period and are calculated following the American Society for Civil Engineers formulation of the Penman–Monteith method driven by seasonal climate forecasts of monthly mean temperature, humidity, wind speed, and solar radiation from the National Centers for Environmental Prediction CFSv2 model and the National Aeronautics and Space Administration GEOS-5 model. The skill evaluation, using deterministic and probabilistic scores, focuses on the December–February (DJF), March–May (MAM), June–August (JJA), and September–November seasons. The results indicate that ET0 forecasts are a promising tool for early warning of drought and food insecurity. Globally, the regions where forecasts are most skillful (correlation > 0.35 at leads of 2 months) include the western United States, northern parts of South America, parts of the Sahel region, and southern Africa. The FEWS NET regions where forecasts are most skillful (correlation > 0.35 at lead 3) include northern sub-Saharan Africa (DJF; dry season), Central America (DJF; dry season), parts of East Africa (JJA; wet season), southern Africa (JJA; dry season), and central Asia (MAM; wet season). A case study over parts of East Africa for the JJA season shows that ET0 forecasts in combination with the precipitation forecasts would have provided early warning of recent severe drought events (e.g., in 2002, 2004, 2009) that contributed to substantial food insecurity in the region.

Flash droughts can be distinguished by rapid intensification from near-normal soil moisture to drought conditions in a matter of weeks. Here, we provide the first characterisation of a climatology of flash drought across Australia using a suite of indices. The experiment is designed to capture a range of conditions related to drought: evaporative demand describes the atmospheric demand for moisture from the surface; precipitation, the supply of moisture from the atmosphere to the surface; and evaporative stress, the supply of moisture from the surface relative to the demand from the atmosphere. We show that regardless of the definition, flash droughts occur in all seasons. They can terminate as rapidly as they start, but in some cases can last many months, resulting in a seasonal-scale drought. We show that flash-drought variability and its prevalence can be related to phases of the El Niño–Southern Oscillation, highlighting scope for seasonal-scale prediction. Using a case study in southeast Australia, we show that monitoring precipitation is less useful for capturing the onset of flash drought as it occurs. Instead, indices like the Evaporative Demand Drought Index and Evaporative Stress Index are more useful for monitoring flash-drought development.

Precipitation, soil moisture, and air temperature are the most commonly used climate variables to monitor drought; however, other climatic factors such as solar radiation, wind speed, and humidity can be important drivers in the depletion of soil moisture and evolution and persistence of drought. This work assesses the Evaporative Demand Drought Index (EDDI) at multiple time scales for several hydroclimates as the second part of a two-part study. EDDI and individual evaporative demand components were examined as they relate to the dynamic evolution of flash drought over the central United States, characterization of hydrologic drought over the western United States, and comparison to commonly used drought metrics of the U.S. Drought Monitor (USDM), Standardized Precipitation Index (SPI), Standardized Soil Moisture Index (SSI), and the evaporative stress index (ESI). Two main advantages of EDDI over other drought indices are that it is independent of precipitation (similar to ESI) and it can be decomposed to identify the role individual evaporative drivers have on drought onset and persistence. At short time scales, spatial distributions and time series results illustrate that EDDI often indicates drought onset well in advance of the USDM, SPI, and SSI. Results illustrate the benefits of physically based evaporative demand estimates and demonstrate EDDI’s utility and effectiveness in an easy-to-implement agricultural early warning and long-term hydrologic drought–monitoring tool with potential applications in seasonal forecasting and fire-weather monitoring.

Many operational drought indices focus primarily on precipitation and temperature when depicting hydroclimatic anomalies, and this perspective can be augmented by analyses and products that reflect the evaporative dynamics of drought. The linkage between atmospheric evaporative demand E0 and actual evapotranspiration (ET) is leveraged in a new drought index based solely on E0—the Evaporative Demand Drought Index (EDDI). EDDI measures the signal of drought through the response of E0 to surface drying anomalies that result from two distinct land surface–atmosphere interactions: 1) a complementary relationship between E0 and ET that develops under moisture limitations at the land surface, leading to ET declining and increasing E0, as in sustained droughts, and 2) parallel ET and E0 increases arising from increased energy availability that lead to surface moisture limitations, as in flash droughts. To calculate EDDI from E0, a long-term, daily reanalysis of reference ET estimated from the American Society of Civil Engineers (ASCE) standardized reference ET equation using radiation and meteorological variables from the North American Land Data Assimilation System phase 2 (NLDAS-2) is used. EDDI is obtained by deriving empirical probabilities of aggregated E0 depths relative to their climatologic means across a user-specific time period and normalizing these probabilities. Positive EDDI values then indicate drier-than-normal conditions and the potential for drought. EDDI is a physically based, multiscalar drought index that that can serve as an indicator of both flash and sustained droughts, in some hydroclimates offering early warning relative to current operational drought indices. The performance of EDDI is assessed against other commonly used drought metrics across CONUS in Part II.

Evaporative demand (E0) both drives and responds to droughts based on interactions across the land surface-atmosphere interface, and can be exploited to signal agricultural, hydrologic, and ecological droughts. In this chapter, we argue that using a fully physically based measure of E0 moves the drought community toward a more complete understanding of drought processes that will enhance its abilities with regard to early warning and drought monitoring in the present day and drought-risk assessment under future climate change scenarios. We examine regional characteristics in E0 and their behavior during droughts in the recent historical period across different hydroclimates. We review physical mechanisms driving extremes in E0 and how climate change could influence their behavior in both driving and responding to extreme droughts in the 21st century. Finally, we discuss plant physiological responses under elevated atmospheric CO2 and their consequences for drought and heat stress on plants.

A novel contiguous United States (CONUS) wide evaluation of reference evapotranspiration (ET0; a formulation of evaporative demand) anomalies is performed using the Climate Forecast System version 2 (CFSv2) reforecast data for 1982–2009. This evaluation was motivated by recent research showing ET0 anomalies can accurately represent drought through exploitation of the complementary relationship between actual evapotranspiration and ET0. Moderate forecast skill of ET0 was found up to leads of 5 months and was consistently better than precipitation skill over most of CONUS. Forecasts of ET0 during drought events revealed high categorical skill for notable warm-season droughts of 1988 and 1999 in the central and northeast CONUS, with precipitation skill being much lower or absent. Increased ET0 skill was found in several climate regions when CFSv2 forecasts were initialized during moderate-to-strong El Nino–Southern Oscillation events. Our findings suggest that ET0 anomaly forecasts can improve and complement existing seasonal drought forecasts.

The National Oceanic and Atmospheric Administration (NOAA) provides daily reference evapotranspiration (ETref) maps for the contiguous United States using climatic data from North American Land Data Assimilation System (NLDAS). This data provides large-scale spatial representation of ETref, which is essential for regional scale water resources management. Data used in the development of NOAA daily ETref maps are derived from observations over surfaces that are different from short (grass — ETos) or tall (alfalfa — ETrs) reference crops, often in nonagricultural settings, which carries an unknown discrepancy between assumed and actual conditions. In this study, NOAA daily ETos and ETrs maps were evaluated for accuracy, using observed data from the Texas High Plains Evapotranspiration (TXHPET) network. Daily ETos, ETrs and the climatic data (air temperature, wind speed, and solar radiation) used for calculating ETref were extracted from the NOAA maps for TXHPET locations and compared against ground measurements on reference grass surfaces. NOAA ETref maps generally overestimated the TXHPET observations (1.4 and 2.2 mm/day ETos and ETrs, respectively), which may be attributed to errors in the NLDAS modeled air temperature and wind speed, to which reference ETref is most sensitive. Therefore, a bias correction to NLDAS modeled air temperature and wind speed data, or adjustment to the resulting NOAA ETref, may be needed to improve the accuracy of NOAA ETref maps.

The potential evapotranspiration (PET) that would occur with unlimited plant access to water is a central driver of simulated plant growth in many ecological models. PET is influenced by solar and longwave radiation, temperature, wind speed, and humidity, but it is often modeled as a function of temperature alone. This approach can cause biases in projections of future climate impacts in part because it confounds theeffects of warming due toincreasedgreenhousegases with thatwhich wouldbe caused by increased radiation from the sun. We developed an algorithm for linking PET to extraterrestrial solar radiation (incoming top-of atmosphere solar radiation), as well as temperature and atmospheric water vapor pressure, and incorporated this algorithm into the dynamic global vegetation model MC1. We tested the newalgorithm for the Northern Great Plains, USA, whose remaining grasslands are threatened by continuing woody encroachment. Both the new and the standard temperature-dependent MC1 algorithm adequately simulated current PET, as compared to the more rigorous PenPan model of Rotstayn et al. (2006). However, compared to the standard algorithm, the new algorithm projected a much more gradual increase in PET over the 21st century for three contrasting future climates. This difference led to lower simulated drought effects and hence greater woody encroachment with the new algorithm, illustrating the importance of more rigorous calculations of PET in ecological models dealing with climate change.

Terrestrial evapotranspiration (ET) is a major component of the surface hydrological cycle and controls land-atmosphere feedbacks by modulating the surface energy budget. Accurate ET quantification at global or regional scales is crucial for understanding variations in carbon and water cycling in a changing environment. Although various grid-based ET data models have been developed using multiple approaches, these vary in concept and physical scheme, leading to differences in performance. We examine uncertainties associated with the limitations of the physics used to assist in model selection and improvement. We evaluate multiple ET data models, including estimates derived from a variety of land surface models (LSMs) based on the operational North American Land Data Assimilation System (NLDAS) phase 2 (NLDAS-2) and the experimental NASA LIS-based NLDAS Testbed (NLDAS-Testbed) drivers, and satellite retrievals, compared to water budget-derived ET and tower observations. Overall, all models are able to capture the spatial variability of mean annual water balance-based ET (ETwb) and monthly seasonal cycles of tower ET measurements, although there is a large range of estimates. NOAH28, FLUXNET, SSEBop, LandFlux, and GLEAM perform best, as demonstrated by their higher correlation and smaller bias and RMSE values. Simple relative uncertainty analysis shows that the NLDAS-Testbed ensemble mean has a slightly lower uncertainty than that of the NLDAS-2 ensemble. Our study indicates that NLDAS-Testbed/VIC412 (NLDAS version/LSM version) is improving and NLDAS-Testbed /CLSM is deteriorating relative to NLDAS-2/VIC403 and NLDAS-2/Mosaic. NLDAS-Testbed /NOAH36 and NLDAS-Testbed /NOAHMP36 are comparable to NLDAS-2/NOAH28, although biases between models and ETwb exhibit opposite trends. These findings will help further improvement of these models and support future NLDAS development.

The North American Land Data Assimilation System project phase 2 (NLDAS-2) has run four land surface models for a 30-year (1979–2008) retrospective period. Land surface evapotranspiration (ET) is one of the most important model outputs from NLDAS-2 for investigating land–atmosphere interaction or to monitor agricultural drought. Here, we evaluate hourly ET using in situ observations over the Southern Great Plains (Atmospheric Radiation Measurement/Cloud and Radiation Testbed network) for 1 January 1997–30 September 1999 and daily ET u-sing in situ observations at the AmeriFlux network over the conterminous USA for an 8-year period (2000–2007). The NLDAS-2 models compare well against observations, with the National Centers for Environmental Prediction's Noah land surface model performing best, followed, in order, by the Variable Infiltration Capacity, Sacramento Soil Moisture Accounting, and Mosaic models. Daily evaluation across the AmeriFlux network shows that for all models, performance depends on season and vegetation type; they do better in spring and fall than in winter or summer and better for deciduous broadleaf forest and grasslands than for croplands or evergreen needleleaf forest. Copyright © 2014 John Wiley & Sons, Ltd.

Concurrent with the trend of rising global average air temperature, there have been worldwide observations of a decline in pan evaporation over the last 30–50 years. This global phenomenon has since received much attention from the scientific community. Better interpretation of the long-term trend of pan evaporation (involving seasonal and inter-annual variations) requires rigorous experimental investigation of the physics of pan evaporation. To do that, we constructed an instrumented US Class A pan that replicated an operational pan at Canberra Airport in Australia. We subsequently monitored pan evaporation and associated meteorological variables at half-hourly intervals over a three-year period (Oct 2007–Jan 2010). Extending our earlier work on the aerodynamics of pan evaporation, we conducted a theoretical and experimental study on the energy balance of the pan under non-steady state conditions. The theory considers heat exchanges at both the pan water surface and the pan wall. We formulate the radiative balance based on the geometry of the system (the pan diameter, the height of the water level, the height of the rim and solar zenith angle) and account for differences in albedo and emissivity between the pan water surface and the pan wall. The theory is used to compute all relevant energy fluxes and thereby close the energy balance. Integration of the half-hourly fluxes to a daily basis showed that we were able to close the energy budget with a RMSE of ∼18 W m −2 or 7.5% of the net short-wave radiation. We find that evaporation from our pan is dominated (∼80%) by the radiative exchanges at the pan water surface; those at the pan wall are smaller but also important. Contrary to expectation, there was little if any sensible heat exchange across the pan wall. Instead we found that the major sensible heat flux occurs across the pan water surface. Importantly, evaporative cooling at the pan water surface can result in a transfer of sensible heat from the (warmer) air to the (cooler) surface of the water body. The sensible heat flux from the overlying air to the pan water surface is minimal at low evaporation rates, but at higher evaporation rates (e.g., >300 W m −2 ) it contributes the equivalent of up to 15% of the total evaporative flux.

Assessing climate change risk to municipal water supplies is often conducted by hydrologic modeling specific to local watersheds and infrastructure to ensure that outputs are compatible with existing planning frameworks and processes. This study leverages the modeling capacity of an operational National Weather Service River Forecast Center to explore the potential impacts of future climate-driven hydrologic changes on factors important to planning at the Salt Lake City Department of Public Utilities (SLC). Hydrologic modeling results for the study area align with prior research in showing that temperature changes alone will lead to earlier runoff and reduced runoff volume. The sensitivity of average annual flow to temperature varies significantly between watersheds, averaging −3.8% °F−1 and ranging from −1.8% to −6.5% flow reduction per degree Fahrenheit of warming. The largest flow reductions occur during the high water demand months of May–September. Precipitation drives hydrologic response more strongly than temperature, with each 1% precipitation change producing an average 1.9% runoff change of the same sign. This paper explores the consequences of climate change for the reliability of SLC's water supply system using scenarios that include hydrologic changes in average conditions, severe drought scenarios, and future water demand test cases. The most significant water management impacts will be earlier and reduced runoff volume, which threaten the system's ability to maintain adequate streamflow and storage to meet late-summer water demands.

To understand the sources of temporal and spatial variability of atmospheric evaporative demand across the conterminous United States (CONUS), a mean-value, second-moment uncertainty analysis is applied to a spatially distributed dataset of daily synthetic pan evaporation for 1980–2009. This evaporative demand measure is from the “PenPan” model, which is a combination equation calibrated to mimic observations from U.S. class-A evaporation pans and here driven by six North American Land Data Assimilation System variables: temperature, specific humidity, station pressure, wind speed, and downwelling shortwave and longwave radiation. The variability of evaporative demand is decomposed across various time scales into contributions from these drivers. Contrary to popular expectation and much hydrologic practice, temperature is not always the most significant driver of temporal variability in evaporative demand, particularly at subannual time scales. Instead, depending on the season, one of four drivers (temperature, specific humidity, downwelling shortwave radiation, and wind speed) dominates across different regions of CONUS. Temperature generally dominates in the northern continental interior. This analysis assists land surface modelers in balancing parameter parsimony and physical representativeness. Patterns of dominant drivers are shown to cycle seasonally, with clear implications for modeling evaporative demand in operational hydrology or as a metric of climate change and variability. Depending on the region and season, temperature, specific humidity, downwelling shortwave radiation, and wind speed must together be examined, with downwelling longwave radiation as a secondary input. If any variable may be ignored, it is atmospheric pressure. Parameterizations of evaporative demand based solely on temperature should be avoided at all time scales.

In response to worldwide observations reporting a decline in pan evaporation over the last 30–50 years, we developed an instrumented US Class A pan that replicated an operational pan at Canberra Airport in Australia. The aim of the experimental facility was to investigate the physics of pan evaporation under non-steady state conditions. By monitoring the water level at 5-min intervals we were able to calculate the evaporation rate and thereby determine the short-term mass balance of the pan. Over the same time intervals, we also monitored (short- and long-wave) radiation, temperature (air, water surface, bulk water, inner and outer pan wall), atmospheric pressure as well as the air vapour pressure and the wind speed at a standard reference height (2 m above ground level). The experimental pan was operated for three years (2007–2010). In this paper, we develop a framework for quantifying vapour transfer by coupling Fick's First Law of Diffusion with boundary layer theory. This approach adequately represented pan evaporation measurements over short time intervals (half-hourly) under non-steady state conditions provided that surface temperature measurements, that account for the substantial cooling associated with evaporation, are available. It involved estimating the boundary layer thickness and other properties of air above the evaporating surface for a pan. Our results are consistent with the “envelope of theoretical curves” concept for the wind function introduced by Thom et al. (1981) .

Pan evaporation is just that - it is the evaporation rate of water from a small dish located at the ground-surface. Pan evaporation is a measure of the evaporative demand over terrestrial surfaces. Declines in pan evaporation have now been reported in many regions of the world. The trends vary from one pan to the next, but when averaged over many pans, they are typically in the range of −1 to − 4m m a −2 (mm per annum per annum). In energetic terms, a trend of − 2m m a −2 is equivalent to −0.16 W m −2 a −1 and over 30 years this is a change of −4.8 W m −2 . For comparison, the top-of-atmosphere forcing due to doubled CO2 is estimated by the Intergovernmental Panel on Climate Change (IPCC) to be ~3.7 W m −2 . Hence, the magnitude of the pan evaporation trend is large. What is of even greater interest is the direction - a decline - given the well-established warming of the last 30-50 years. In this article, the first in a two part series, we describe the underlying principles in using and interpreting pan evaporation data and then summarise the reported observations from different countries. In the second article, we describe the interpretation of the trends in terms of changes in the terrestrial water balance.

Available water supply across the contiguous 48 states was estimated as precipitation minus evapotranspiration using data for the period 1953-1994. Precipitation estimates were taken from the Parameter-Elevation Regressions on Independent Slopes Model (PRISM). Evapotranspiration was estimated using two models, the Advection-Aridity model and the Zhang model. The evapotranspiration models were calibrated using precipitation and runoff data for 655 hydrologically undisturbed basins, and then tested using estimates of natural runoff for the 18 water resource regions (WRR) of the 48 contiguous states. The final water supply coverage reflects a mixture of outputs from the two evapotranspiration models. Political, administrative, and land cover boundaries were mapped over the coverage of mean annual water supply. Across the entire study area, we find that 53% of the water supply originates on forested land, which covers only 29% of the surface area, and that 24% originates on federal lands, including 18% on national forests and grasslands alone. Forests and federal lands are even more important in the West (the 11 western contiguous states), where 65% of the water supply originates on forested land and 66% on federal lands, with national forests and grasslands contributing 51%.

Several studies have projected increases in drought severity, extent and duration in many parts of the world under climate change. We examine sources of uncertainty arising from the methodological choices for the assessment of future drought risk in the continental US (CONUS). One such uncertainty is in the climate models' expression of evaporative demand (E0), which is not a direct climate model output but has been traditionally estimated using several different formulations. Here we analyze daily output from two CMIP5 GCMs to evaluate how differences in E0 formulation, treatment of meteorological driving data, choice of GCM, and standardization of time series influence the estimation of E0. These methodological choices yield different assessments of spatio-temporal variability in E0 and different trends in 21st century drought risk. First, we estimate E0 using three widely used E0 formulations: Penman-Monteith; Hargreaves-Samani; and Priestley-Taylor. Our analysis, which primarily focuses on the May-September warm-season period, shows that E0 climatology and its spatial pattern differ substantially between these three formulations. Overall, we find higher magnitudes of E0 and its interannual variability using Penman-Monteith, in particular for regions like the Great Plains and southwestern US where E0 is strongly influenced by variations in wind and relative humidity. When examining projected changes in E0 during the 21st century, there are also large differences among the three formulations, particularly the Penman-Monteith relative to the other two formulations. The 21st century E0 trends, particularly in percent change and standardized anomalies of E0, are found to be sensitive to the long-term mean value and the amplitude of interannual variability, i.e. if the magnitude of E0 and its interannual variability are relatively low for a particular E0 formulation, then the normalized or standardized 21st century trend based on that formulation is amplified relative to other formulations. This is the case for the use of Hargreaves-Samani and Priestley-Taylor, where future E0 trends are comparatively much larger than for Penman-Monteith. When comparing Penman-Monteith E0 responses between different choices of input variables related to wind speed, surface roughness, and net radiation, we found differences in E0 trends, although these choices had a much smaller influence on E0 trends than did the E0 formulation choices. These methodological choices and specific climate model selection, also have a large influence on the estimation of trends in standardized drought indices used for drought assessment operationally. We find that standardization tends to amplify divergences between the E0 trends calculated using different E0 formulations, because standardization is sensitive to both the climatology and amplitude of interannual variability of E0. For different methodological choices and GCM output considered in estimating E0, we examine potential sources of uncertainty in 21st century trends in the Standardized Precipitation Evapotranspiration Index (SPEI) and Evaporative Demand Drought Index (EDDI) over selected regions of the CONUS to demonstrate the practical implications of these methodological choices for the quantification of drought risk under climate change.

Two implementations of the complementary relationship hypothesis for regional evapotranspiration, the Complementary Relationship Areal Evapotranspiration (CRAE) model and the Advection-Aridity (AA) model, are evaluated against independent estimates of regional evapotranspiration derived from long-term, large-scale water balances (1962-1988) for 120 minimally impacted basins in the conterminous United States. The CRAE model overestimates annual evapotranspiration by 2.5% of mean annual precipitation, and the AA model underestimates annual evapotranspiration by 10.6% of precipitation. Generally, increasing humidity leads to decreasing absolute errors for both models, and increasing aridity leads to increasing overestimation by the CRAE model and underestimation by the AA model, with the exception of high, arid basins, where the AA model overestimates evapotranspiration. Overall, the results indicate that the advective portion of the AA model must be recalibrated before it may be used successfully on a regional basis and that the CRAE model accurately predicts monthly regional evapotranspiration.

Long-term monthly evapotranspiration estimates from Brutsaert and Stricker's Advection-Aridity model were compared with independent estimates of evapotranspiration derived from long-term water balances for 139 undisturbed basins across the conterminous United States. On an average annual basis for the period 1962-1988 the original model, which uses a Penman wind function, underestimated evapotranspiration by 7.9% of precipitation compared with the water balance estimates. Model accuracy increased with basin humidity. An improved formulation of the model is presented in which the wind function and the Priestley-Taylor coefficient are modified. The wind function was reparameterized on a seasonal, regional basis to replicate independent proxy potential evapotranspiration surfaces. This led to significant differences from the original Penman wind function. The reparameterized wind function, together with a recalibrated Priestley- Taylor coefficient in the wet environment evapotranspiration formulation, reduced the underestimation of annual average evapotranspiration to only 1.15% of precipitation on an independent set of validation basins. The results offered here lend further support for Bouchet's hypothesis as it applies to large-scale, long-term evapotranspiration.

[1] We examine the effects of two different parameterizations of potential evaporation on long-term trends in soil moisture, evaporative flux and runoff simulated by the water balance model underlying the Palmer Drought Severity Index. The first, traditional parameterization is based on air temperature alone. The second parameterization is derived from observations of evaporation from class-A pans. Trends in potential evaporation from the two parameterizations are opposite in sign (±) at almost half the stations tested over Australia and New Zealand. The sign of trends in the modelled soil moisture, evaporative flux and runoff depends on the parameterization used and on the prevailing climatic regime: trends in water-limited regions are driven by precipitation trends, but the choice of parameterization for potential evaporation is shown to be critical in energy-limited regions.

[1] Evaporative demand, measured by pan evaporation, has declined in many regions over the last several decades. It is important to understand why. Here we use a generic physical model based on mass and energy balances to attribute pan evaporation changes to changes in radiation, temperature, humidity and wind speed. We tested the approach at 41 Australian sites for the period 1975–2004. Changes in temperature and humidity regimes were generally too small to impact pan evaporation rates. The observed decreases in pan evaporation were mostly due to decreasing wind speed with some regional contributions from decreasing solar irradiance. Decreasing wind speeds of similar magnitude has been reported in the United States, China, the Tibetan Plateau and elsewhere. The pan evaporation record is invaluable in unraveling the aerodynamic and radiative drivers of the hydrologic cycle, and the attribution approach described here can be used for that purpose.

Using independent observations of actual and potential evapotranspiration at a wide range of spatial scales, we provide direct observational evidence of the complementary relationship in regional evapotranspiration hypothesized by Bouchet in 1963. Bouchet proposed that, for large homogeneous surfaces with minimal advection of heat and moisture, potential and actual evapotranspiration depend on each other in a complementary manner through land-atmosphere feedbacks. Although much work has been done that has led to important theoretical and conceptual insights about regional actual evapotranspiration and its relation to regional potential evapotranspiration, never before has a data set of direct observations been assembled that so clearly displays complementarity, providing strong evidence for the complementary relationship hypothesis, and raising its status above that of a mere conjecture.

[1] Pan evaporation (ETpan) has decreased at 64% of pans in the conterminous U.S. over the past half-century. Comparing trends in ETpan and water budget-derived actual evapotranspiration (ET*a), we observe the so-called “Pan Evaporation Paradox,” which we confirm is no more than a manifestation of the complementarity between actual evapotranspiration (ETa) and potential evapotranspiration (ETp). Examining trends in the components of ETa—the radiative energy and regional advective budgets—we show that both components must be considered together to explain the relationship between ETpan and ET*a.