Search

A recent theory proposes that inland continental regions are in a state of surface flux equilibrium (SFE), in which tight coupling between the land and atmosphere allow estimation of the Bowen ratio at daily to monthly time scales solely from atmospheric measurements, without calibration, even when the land surface strongly constrains the surface energy budget. However, since the theory has only been evaluated at quasi-point spatial scales using eddy covariance measurements with limited global coverage, it is unclear if it is applicable to the larger spatial scales relevant to studies of global climate. In this study, SFE estimates of the Bowen ratio are combined with satellite observations of surface net radiation to obtain large-scale estimates of latent heat flux λE. When evaluated against multiyear mean annual λE obtained from catchment water balance estimates from 221 catchments across the United States, the resulting error statistics are comparable to those in the catchment water balance estimates themselves. The theory is then used to diagnostically estimate λE using historical simulations from 26 CMIP6 models. The resulting SFE estimates are typically at least as accurate as the CMIP6 model's simulated λE, when compared with catchment water balance estimates. Globally, there is broad spatial and temporal agreement between CMIP6 model SFE estimates and the CMIP6 model's simulated λE, although SFE likely overestimates λE in some arid regions. We conclude that SFE applies reasonably at large spatial scales relevant to climate studies, and is broadly reproduced in climate models. [ABSTRACT FROM AUTHOR]

The coupling of soil moisture (SM) and evapotranspiration (ET) is a critical process of the terrestrial climate and water cycle, whose simulation in climate models exhibits substantial uncertainties. Here we investigate, across phase 5 of the Coupled Model Intercomparison Project models in present‐day and future simulations, how this coupling manifests itself across the different components of ET: soil evaporation, transpiration, and canopy interception. We characterize summertime SM‐ET coupling by (interannual) correlations, which we decompose into terms attributable to each ET component. The transpiration and soil evaporation terms share similar spatial patterns, but the contribution of transpiration, globally, is less positive. Canopy interception contributes a positive term to SM‐ET coupling, reflecting the noncausal, rainfall‐forced positive correlation between SM and canopy interception. Model differences are greatest for the transpiration term, which explains most of the model spread in SM‐ET coupling. Models project a robust pattern of more positive SM‐ET correlations in the future. In parts of the midlatitudes and Tropics, this increase reflects reduced precipitation and increased SM limitation on transpiration and soil evaporation. However, at higher latitudes (north of 50°N), increased SM‐ET coupling is driven by the increased contribution of canopy interception induced by the increase in vegetation and precipitation. Analysis of ET partitioning is thus essential to the interpretation of simulated changes in ET and its drivers: While increased SM‐ET correlations may suggest a widespread increase in SM limitation on ET in a warmer world, increases in actual SM control on land‐atmosphere water fluxes are generally limited to regions of negative precipitation change. Key Points: We decompose soil moisture (SM)‐evapotranspiration (ET) coupling into terms attributable to the different ET componentsPart of SM‐ET coupling reflects the noncausal, positive correlation of soil moisture with canopy interceptionProjected increases in SM‐ET coupling in regions of precipitation increase reflect the greater role of canopy interception [ABSTRACT FROM AUTHOR]

Evapotranspiration (ET) is a key process affecting terrestrial hydroclimate, as it modulates the land surface carbon, energy, and water budgets. Evapotranspiration mainly consists of the sum of three components: plant transpiration, soil evaporation, and canopy interception. Here we investigate how the partitioning of ET into these three main components is represented in CMIP5 model simulations of present and future climate. A large spread exists between models in the simulated mean present-day partitioning; even the ranking of the different components in the global mean differs between models. Differences in the simulation of the vegetation leaf area index appear to be an important cause of this spread. Although ET partitioning is not accurately known globally, existing global estimates suggest that CMIP5 models generally underestimate the relative contribution of transpiration. Differences in ET partitioning lead to differences in climate characteristics over land, such as land–atmosphere fluxes and near-surface air temperature. On the other hand, CMIP5 models simulate robust patterns of future changes in ET partitioning under global warming, notably a marked contrast between decreased transpiration and increased soil evaporation in the tropics, whereas transpiration and evaporation both increase at higher latitudes and both decrease in the dry subtropics. Idealized CMIP5 simulations from a subset of models show that the decrease in transpiration in the tropics largely reflects the stomatal closure effect of increased atmospheric CO2 on plants (despite increased vegetation from CO2 fertilization), whereas changes at higher latitudes are dominated by radiative CO2 effects, with warming and increased precipitation leading to vegetation increase and simultaneous (absolute) increases in all three ET components. [ABSTRACT FROM AUTHOR]

Soilmoisture-atmosphere coupling is a key process underlying climate variability and change over land. The control of soil moisture (SM) on evapotranspiration (ET) is a necessary condition for soil moisture to feed back onto surface climate. Here we investigate how this control manifests itself across simulations from the CMIP5 ensemble, using correlation analysis focusing on the interannual (summertime) time scale. Analysis of CMIP5 historical simulations indicates significantmodel diversity in SM-ET coupling in terms of patterns and magnitude. We investigate the relationship of this spread with differences in background simulated climate. Mean precipitation is found to be an important driver of model spread in SM-ET coupling but does not explain all of the differences, presumably because ofmodel differences in the treatment of land hydrology. Compared to observations, some land regions appear consistently biased dry and thus likely overly soil moisture-limited. Because of ET feedbacks on air temperature, differences in SM-ET coupling induce model uncertainties across the CMIP5 ensemble in mean surface temperature and variability.We explore the relationships between model uncertainties in SM-ET coupling and climate projections. In particular over mid-to-high-latitude continental regions of the Northern Hemisphere but also in parts of the tropics, models that are more soilmoisture-limited in the present tend to warm more in future projections, because they project less increase in ET and (in midlatitudes) greater increase in incoming solar radiation. Soil moisture-atmosphere processes thus contribute to the relationship observed across models between summertime present-day simulated climate and future warming projections over land. [ABSTRACT FROM AUTHOR]