Soil Moisture Precipitation Coupling Research Articles

Causal deep learning models for studying the Earth system

Abstract. Earth is a complex non-linear dynamical system. Despite decades of research and considerable scientific and methodological progress, many processes and relations between Earth system variables remain poorly understood. Current approaches for studying relations in the Earth system rely either on numerical simulations or statistical approaches. However, there are several inherent limitations to existing approaches, including high computational costs, uncertainties in numerical models, strong assumptions about linearity or locality, and the fallacy of correlation and causality. Here, we propose a novel methodology combining deep learning (DL) and principles of causality research in an attempt to overcome these limitations. On the one hand, we employ the recent idea of training and analyzing DL models to gain new scientific insights into relations between input and target variables. On the other hand, we use the fact that a statistical model learns the causal effect of an input variable on a target variable if suitable additional input variables are included. As an illustrative example, we apply the methodology to study soil-moisture–precipitation coupling in ERA5 climate reanalysis data across Europe. We demonstrate that, harnessing the great power and flexibility of DL models, the proposed methodology may yield new scientific insights into complex non-linear and non-local coupling mechanisms in the Earth system.

A sensitivity study on the response of convection initiation to in situ soil moisture in the central United States

Soil moisture can affect precipitation, however there is significant debate about the sign and strength of land–atmosphere coupling. This may be due to the differences regarding the selection of data sources, the presence of atmospheric persistence, and the spatial and temporal scales of the analysis. This study is a sensitivity analysis that quantifies how each of these factors influences land–atmosphere coupling in the central United States from 2005 to 2007. Four different sources of soil moisture are used to represent soil moisture: in situ measurements and three satellite products. The Thunderstorm Observation by Radar (ThOR) algorithm is used to identify convective events. The results show that warm-season afternoon convection initiates preferentially over dry soil, if in situ measurements are used. However, when satellite data are used the sign and strength of the soil moisture-precipitation coupling varies depending on which satellite dataset is used. This demonstrates that evaluations of land–atmosphere coupling strength are sensitive to the source of soil moisture. Our results also show that accounting for atmospheric persistence tends to weaken apparent land–atmosphere coupling. Changing the spatial scale of the analysis can also influence the sign and coupling strength. Analyses that are performed at a coarser spatial resolution tend to indicate a wet preference because they do not capture the local negative feedbacks. This study demonstrates that contradictions in the literature regarding the sign and strength of soil moisture-precipitation feedbacks can be partly attributed to differences in datasets, methods and scale of the analysis.

Soil moisture–precipitation coupling over Central Europe: Interactions between surface anomalies at different scales and the dynamical implication

The sign of soil moisture–precipitation coupling and its scale dependence are controversial issues when examining soil–atmosphere interactions. Using the operational convection‐permitting COSMO model with 2.8 km horizontal grid spacing, the interplay of soil‐moisture heterogeneities on different scales and the role of the midtropospheric background wind are addressed by performing a series of real‐case simulations for 11 synoptically weakly and moderately forced summer precipitation events across Central Europe. The cumulative effect of soil‐moisture bias and heterogeneities on different scales is investigated, with consideration of the prevailing synoptic forcing, by explicitly perturbing soil‐moisture conditions. A soil‐moisture bias of ±25% is combined with different soil‐moisture heterogeneity length‐scales ranging from 30 to 140 km introduced by chessboard patterns.For synoptically moderately forced cases, our experiments only show a small impact of soil‐moisture perturbations on convective afternoon precipitation. Averaged over all seven synoptically weakly forced cases, however, we find (a) a positive coupling between the overall soil‐moisture bias and the domain‐averaged precipitation and (b) a negative local soil moisture–precipitation coupling. Increased precipitation over drier soils is related to an interaction between thermally induced vertical circulations and the background wind, causing a persistent updraft region at the downstream flank of the dry patches. Convective precipitation is triggered preferentially near these soil‐moisture gradients. The circulation cells are most dominant for soil‐moisture perturbations at scales between 40 and 80 km. The spatial locking of convection at soil‐moisture boundaries at these scales results in an earlier triggering of convection and an overall reduction of the day‐to‐day variability of area‐averaged precipitation.

Regional co-variability of spatial and temporal soil moisture–precipitation coupling in North Africa: an observational perspective

Abstract. The magnitude and sign of soil moisture–precipitation coupling (SMPC) is investigated using a probability-based approach and 10 years of daily microwave satellite data across North Africa at a 1∘ horizontal scale. Specifically, the co-existence and co-variability of spatial (i.e. using soil moisture gradients) and temporal (i.e. using soil moisture anomaly) soil moisture effects on afternoon rainfall is explored. The analysis shows that in the semi-arid environment of the Sahel, the negative spatial and the negative temporal coupling relationships do not only co-exist, but are also dependent on one another. Hence, if afternoon rain falls over temporally drier soils, it is likely to be surrounded by a wetter environment. Two regions are identified as SMPC “hot spots”. These are the south-western part of the domain (7–15∘ N, 10∘ W–7∘ E), with the most robust negative SMPC signal, and the South Sudanese region (5–13∘ N, 24–34∘ E). The sign and significance of the coupling in the latter region is found to be largely modulated by the presence of wetlands and is susceptible to the number of long-lived propagating convective systems. The presence of wetlands and an irrigated land area is found to account for about 30 % of strong and significant spatial SMPC in the North African domain. This study provides the first insight into regional variability of SMPC in North Africa, and supports the potential relevance of mechanisms associated with enhanced sensible heat flux and mesoscale variability in surface soil moisture for deep convection development.

Multiscale Land–Atmosphere Coupling and Its Application in Assessing Subseasonal Forecasts over East Asia

Abstract The land surface, with a memory longer than the atmosphere in nature, has been recognized as an important source for Subseasonal to Seasonal (S2S) predictability through land–atmosphere coupling at multiple time scales. Understanding of the land–atmosphere coupling is important for improving subseasonal forecasting that is expected to fill the gap between medium-range weather forecasts and seasonal forecasts. Based on reanalysis and S2S reforecast datasets, land–atmosphere coupling is investigated over East Asia from daily to monthly time scales during summertime. Reanalysis results show that soil moisture–evapotranspiration (ET) coupling is closely related to the monsoonal rain belt shift. The coupling can be significant over humid regions (e.g., south China) during postmonsoon periods, where soil is usually drier, but insignificant over semiarid regions (e.g., north China) after the arrival of a monsoon, where soil is wetter. The dependence of soil moisture–ET coupling on soil wetness conditions decreases as the time scale increases, indicating more significant coupling at longer time scales. Similar sensitivities to time scales are found between ET and lifting condensation level (LCL), and between ET and precipitation, especially over land–atmosphere coupling hotspots. Monthly coupling strength analysis shows that ET–LCL coupling is a key process that determines the soil moisture–precipitation coupling, and the response of convective instability to ET is stronger at longer time scales. Subseasonal forecasting models also show more significant land–atmosphere coupling at monthly than daily time scales, where the ECMWF and NCEP models that best reproduce the coupling and its changes with monsoonal rain belt shifts have the best precipitation forecast skill among S2S models.

Effect of Soil Moisture on Diurnal Convection and Precipitation in Large-Eddy Simulations

Abstract A determination of the sign and magnitude of the soil moisture–precipitation feedback relies either on observations, where synoptic variability is difficult to isolate, or on model simulations, which suffer from biases mainly related to poorly resolved convection. In this study, a large-eddy simulation model with a resolution of 250 m is coupled to a land surface model and several idealized experiments mimicking the full diurnal cycle of convection are performed, starting from different spatially homogeneous soil moisture conditions. The goal is to determine under which conditions drier soils may produce more precipitation than wetter ones. The methodology of previous conceptual studies that have quantified the likelihood of convection to be triggered over wet or dry soils is followed but includes the production of precipitation. Although convection can be triggered earlier over dry soils than over wet soils under certain atmospheric conditions, total precipitation is found to always decrease over dry soils. By splitting the total precipitation into its magnitude and duration component, it is found that the magnitude strongly correlates with surface latent heat flux, hence implying a wet soil advantage. Because of this strong scaling, changes in precipitation duration caused by differences in convection triggering are not able to overcompensate for the lack of evaporation over dry soils. These results are further validated using two additional atmospheric soundings and a series of perturbed experiments that consider cloud radiative effects, as well as the effect of large-scale forcing, winds, and plants on the soil moisture–precipitation coupling.

Irrigation Effects on Land–Atmosphere Coupling Strength in the United States

Abstract Land–atmosphere coupling strength describes the degree to which the atmosphere responds (e.g., via changes in precipitation) to changes in the land surface state (e.g., soil moisture). The Midwest and Great Plains of the United States have been shown to be “hot spots” of coupling by many climate models and some observations. However, very few of the modeling studies have reported whether the climate models applied irrigation in the Midwest and Great Plains, where 24%–27% of farmland is irrigated, leaving open the question of whether irrigation affects current estimates of coupling strength. This study used a regional climate model that incorporated dynamic crop growth and precision irrigation (WRF3.3–CLM4crop) to investigate irrigation effects on land–atmosphere coupling strength. Coupling strength was quantified using multiple indices and the irrigated land-induced precipitation was tracked using a back trajectory method. The indices showed a consistent and significant decline in local coupling strength with irrigation in the Midwest and northern Great Plains. These reductions were due to increased soil moisture but decreased local precipitation and lower sensitivity of latent heat flux to soil moisture over irrigated regions. The back trajectories of water vapor transport confirmed that irrigation largely did not contribute to local precipitation. Water vapor from irrigated land was transported to the Midwest and U.S. Northeast where it fell as precipitation, suggesting that irrigation has a broader spatial impact on soil moisture–precipitation coupling than simply through local soil moisture–evapotranspiration coupling. The present study suggests that climate models without irrigation schemes may overestimate the land–atmosphere coupling strength over irrigated agricultural regions but underestimate coupling strength over neighboring nonirrigated regions.

Soil moisture–precipitation coupling: observations from the Oklahoma Mesonet and underlying physical mechanisms

Abstract. Interactions between soil moisture and the atmosphere are driven by the partitioning of sensible and latent heating, through which soil moisture has been connected to atmospheric modifications that could potentially lead to the initiation of convective precipitation. The majority of previous studies linking the land surface to subsequent precipitation have used atmospheric reanalysis or model data sets. In this study, we link in situ observations of soil moisture from more than 100 stations in Oklahoma to subsequent unorganized afternoon convective precipitation. We use hourly next generation (NEXRAD) radar-derived precipitation to identify convective events, and then compare the location of precipitation initiation to underlying soil moisture anomalies in the morning. Overall we find a statistically significant preference for convective precipitation initiation over drier than normal soils, with over 70 % of events initiating over soil moisture below the long-term median. The significant preference for precipitation initiation over drier than normal soils is in contrast with previous studies using satellite-based precipitation to identify the region of maximum precipitation accumulation. We evaluated 19 convective events occurring near Lamont, Oklahoma, where soundings of the atmospheric profile at 06:00 and 12:00 LST are also available. For these events, soil moisture has strong negative correlations with the level of free convection (LFC), planetary boundary layer (PBL) height, and surface temperature changes between 06:00 and 12:00 LST. We also find strong positive correlations between morning soil moisture and morning-to-afternoon changes in convective available potential energy and convective inhibition. In general, the results of this study demonstrate that both positive and negative soil moisture feedbacks are important in this region of the USA.

Contributions of soil moisture interactions to climate change in the tropics in the GLACE–CMIP5 experiment

Contributions of changes in soil moisture to the projected climate change in the tropics at the end of the twenty first century are quantified using the simulations from five different global climate models, which contributed to the GLACE–CMIP5 experiment. “GLACE” refers to the Global Land Atmosphere Coupling Experiment and “CMIP5” to the fifth phase of the Coupled Model Intercomparison Project. This is done by relating the overall projected changes in climate to those changes in climate that are related to the projected changes in soil moisture. The study focusses on two particular aspects of the interactions of the soil moisture with climate, the soil moisture–temperature coupling and the soil moisture–precipitation coupling. The simulations show distinct future changes in soil moisture content in the tropics, with a general tendency of increases in the central parts of the tropics and decreases in the subtropics. These changes are associated with corresponding changes in precipitation, with an overall tendency of an approximate 5 % change in soil moisture in response to a precipitation change of 1 mm/day. All five individual models are characterized by the same qualitative behaviour, despite differences in the strength and in the robustness of the coupling between soil moisture and precipitation. The changes in soil moisture content are found to give important contributions to the overall climate change in the tropics. This is in particularly the case for latent and sensible heat flux, for which about 80 % of the overall changes are related to soil moisture changes. Similarly, about 80 % of the overall near-surface temperature changes (with the mean temperature changes in the tropics removed) are associated with soil moisture changes. For precipitation, on the other hand, about 30–40 % of the overall change can be attributed to soil moisture changes. The robustness of the contributions of the soil moisture changes to the overall climate change varies between the different meteorological variables, with a high degree of robustness for the surface energy fluxes, a fair degree for near-surface temperature and a low degree for precipitation. Similar to the coupling between soil moisture and precipitation, the five individual models are characterized by the same qualitative behaviour, albeit differences in the strength and the robustness of the contributions of the soil moisture change. This suggests that despite the regional differences in the projected climate changes between the individual models, the basic physical mechanisms governing the soil moisture–temperature coupling and the soil moisture–precipitation coupling work similarly in these models. The experiment confirms the conceptual models of the soil moisture–temperature coupling and the soil moisture–precipitation coupling described Seneviratne et al. (Earth-Sci Rev 99:125–161, 2010). For the soil moisture–temperature coupling, decreases (increases) in soil moisture lead to increasing (decreasing) sensible heat fluxes and near-surface temperatures. The soil moisture–precipitation coupling is part of a positive feedback loop, where increases (decreases) in precipitation cause increases (decreases) in soil moisture content, which, in turn, lead to increasing (decreasing) latent heat fluxes and precipitation.

An estimation of the land‐atmosphere coupling strength in South America using the Global Land Data Assimilation System

ABSTRACTThe aim of this study is to identify regions of strong land surface –atmosphere coupling for the austral summer over South America. To accomplish this, a statistical methodology is applied to estimate the interactions of soil moisture with evapotranspiration and precipitation derived from the Global Land Data Assimilation System (GLDAS) dataset. Possible impacts of El Niño Southern Oscillation (ENSO) on the coupling strength are also examined. Particular emphasis is set over two sub‐regions of interest: Southeastern South America (SESA) and the continental part of the South Atlantic Convergence Zone (SACZ). Positive and significant soil moisture–precipitation feedbacks are found over parts of SACZ and in the southern part of South America. Instead, significant negative feedback is found over SESA. The influence of ENSO over the soil moisture–precipitation coupling strength signal is evident over tropical regions. Plausible physical mechanisms involved in the land surface–atmosphere interactions, the influence of ENSO and that of precipitation persistence over extratropical regions on the results, are discussed. The implications of this analysis on monthly to seasonal forecast are also examined. Despite that this methodology cannot be used to establish a precise causal–effect relationship, this study gives a valuable first order approximation of land surface–atmosphere interactions over South America that complements pre‐existing work.

Summer Land–Atmosphere Coupling Strength over the United States: Results from the Regional Climate Model RegCM4–CLM3.5

Abstract This study investigates the land–atmosphere coupling strength during summer over the United States using the Regional Climate Model version 4 (RegCM4)–Community Land Model version 3.5 (CLM3.5). First, a 10-yr simulation driven with reanalysis lateral boundary conditions (LBCs) is conducted to evaluate the model performance. The model is then used to quantify the land–atmosphere coupling strength, predictability, and added forecast skill (for precipitation and 2-m air temperature) attributed to realistic land surface initialization following the Global Land–Atmosphere Coupling Experiment (GLACE) approaches. Similar to previous GLACE results using global climate models (GCMs), GLACE-type experiments with RegCM4 identify the central United States as a region of strong land–atmosphere coupling, with soil moisture–temperature coupling being stronger than soil moisture–precipitation coupling, and confirm that realistic soil moisture initialization is more promising in improving temperature forecasts than precipitation forecasts. At a 1–15-day lead, the added forecast skill reflects predictability (or land–atmosphere coupling strength) indicating that that model can capture the realistic land–atmosphere coupling at a short time scale. However, at a 16–30-day lead, predictability cannot translate to added forecast skill, implying that the coupling at the longer time scale may not be represented well in the model. In addition, comparison of results from GLACE2-type experiments with RegCM4 driven by reanalysis LBCs and those driven by GCM LBCs suggest that the intrinsic land–atmosphere coupling strength within the regional model is the dominant factor for the added forecast skill at a 1–15-day lead, while the impact of LBCs from the GCM may play a dominant role in determining the signal of added forecast skill in the regional model at a 16–30-day lead. It demonstrates the complexities of using regional climate model for GLACE-type studies.

Summer Land–Atmosphere Coupling Strength in the United States: Comparison among Observations, Reanalysis Data, and Numerical Models

Abstract This study examines the land–atmosphere coupling strength during summer over subregions of the United States based on observations [Climate Prediction Center (CPC)–Variable Infiltration Capacity (VIC)], reanalysis data [North American Regional Reanalysis (NARR) and NCEP Climate Forecast System Reanalysis (CFSR)], and models [Community Atmosphere Model, version 3 (CAM3)–Community Land Model, version 3 (CLM3) and CAM4–CLM4]. The probability density function of conditioned correlation between soil moisture and subsequent precipitation or surface temperature during the years of large precipitation anomalies is used as a measure for the coupling strength. There are three major findings: 1) among the eight subregions (classified by land cover types), the transition zone Great Plains (and, to a lesser extent, the Midwest and Southeast) are identified as hot spots for strong land–atmosphere coupling; 2) soil moisture–precipitation coupling is weaker than soil moisture–surface temperature coupling; and 3) the coupling strength is stronger in observational and reanalysis products than in the models examined, especially in CAM4–CLM4. The conditioned correlation analysis also indicates that the coupling strength in CAM4–CLM4 is weaker than in CAM3–CLM3, which is further supported by Global Land–Atmosphere Coupling Experiments1 (GLACE1)-type experiments and attributed to changes in CAM rather than modifications in CLM. Contrary to suggestions in previous studies, CAM–CLM models do not seem to overestimate the land–atmosphere coupling strength.

Land-atmosphere coupling and summer climate variability over East Asia

[1] Two long-term simulations with the weather research and forecasting model are conducted to assess the contribution of land-atmosphere coupling to interannual variability of summer climate over East Asia. The control experiment (CTL) uses a fully coupled land surface model, while an additional experiment replaces soil moisture evolution at each time step with the climatology of CTL and thus removes the interannual variability of soil moisture. CTL is able to reproduce relatively well climatic means and interannual variability of summer climate over East Asia though some biases exist. It is found that land-atmosphere coupling plays a critical role in influencing summer climate variability, in particular over the climatic and ecological transition zones. Interactive soil moisture strongly amplifies daily mean temperature variability over the southern Siberia–northern Mongolia region, the region from northeast China to central China, and the eastern part of South Asia, accounting for half or more of the total variance. Soil moisture is found to exert substantially stronger impacts on daily maximum temperature variability than on daily mean temperature variability but generally has small effects on daily minimum temperature except for the eastern Tibetan Plateau and some other areas. Soil moisture makes a dominant contribution to precipitation variability over the climatic and ecological transition zones of the southern Siberia–northern Mongolia region and northern China and many areas of western China. While soil moisture-temperature coupling is largely determined by the ability of soil moisture to affect surface fluxes, soil moisture–precipitation coupling also depends on other physical processes, particularly moisture convection.

Summer soil—precipitation coupling in South America

The soil moisture memory contributes to atmospheric variability and seasonal predictability and could potentially affect the development of the South American Monsoon System. The relative importance of the local land surface feedbacks and the large-scale dynamical processes during the different phases of the monsoon are still largely unknown. We examine the impacts of land surface conditions during the mature monsoon phase with the Rossby Centre Atmospheric regional model through calculating the coupling strength between soil moisture, evapotranspiration and precipitation. Regions of high coupling strength (hotspots) are identified and analysed focusing on the link between soil moisture—evapotranspiration coupling and soil moisture—precipitation coupling, the relation between the coupling strength and seasonal predictability and the hotspots importance for extreme precipitation events. La Plata Basin and northeastern Brazil are identified as hotspots due to evapotranspiration recycling. A region within the South Atlantic Convergence Zone is identified as a hotspot of precipitation explained by moisture advection. Extreme precipitation events are repressed in parts of La Plata Basin when the link between precipitation and soil moisture is cut through using prescribed soil moisture.

Quantifying the Strength of Soil Moisture–Precipitation Coupling and Its Sensitivity to Changes in Surface Water Budget

Abstract This paper presents a new index to quantify the strength of soil moisture–precipitation coupling in AGCMs and explores how the soil moisture–precipitation coupling in Community Atmosphere Model version 3 (CAM3)–Community Land Model version 3 (CAM3–CLM3) responds to parameterization-induced surface water budget changes. Specifically, this study (a) compares the regions of strong coupling identified by the newly proposed index and the index currently used in the Global Land–Atmosphere Coupling Experiment (GLACE); (b) examines how the surface water budget changes influence the strength of soil moisture–precipitation coupling as measured by the two indexes, respectively; and (c) examines how these changes influence the memory of the coupled land–atmosphere system as measured by the correlation between soil moisture and subsequent precipitation. The new index and the GLACE index are consistent in identifying central North America and West Africa as major regions of strong coupling during June–August (JJA). However, in some areas of western Europe and of subtropical South America where the GLACE index is low, the new index suggests a modest significant coupling during JJA. In response to the surface water budget changes that presumably favor a stronger soil moisture–precipitation coupling, the new index increases, but the GLACE index decreases in a majority of the regions of modest-to-strong coupling, although both show some mixed response. Changes in the land–atmosphere system memory suggest an increase of coupling strength, consistent with results from the new index. The strong dependence of the GLACE index on the relative importance of atmospheric internal variability is identified as a potential cause for the differences between the two indexes. The two indexes emphasize different aspects of soil moisture–precipitation coupling, and one might be more suitable than the other depending on the purpose of individual studies.

Feedbacks of Vegetation on Summertime Climate Variability over the North American Grasslands. Part I: Statistical Analysis

Abstract Feedbacks of vegetation on summertime climate variability over the North American Grasslands are analyzed using the statistical technique of Granger causality. Results indicate that normalized difference vegetation index (NDVI) anomalies early in the growing season have a statistically measurable effect on precipitation and surface temperature later in summer. In particular, higher means and/or decreasing trends of NDVI anomalies tend to be followed by lower rainfall but higher temperatures during July through September. These results suggest that initially enhanced vegetation may deplete soil moisture faster than normal and thereby induce drier and warmer climate anomalies via the strong soil moisture–precipitation coupling in these regions. Consistent with this soil moisture–precipitation feedback mechanism, interactions between temperature and precipitation anomalies in this region indicate that moister and cooler conditions are also related to increases in precipitation during the preceding months. Because vegetation responds to soil moisture variations, interactions between vegetation and precipitation generate oscillations in NDVI anomalies at growing season time scales, which are identified in the temporal and the spectral characteristics of the precipitation–NDVI system. Spectral analysis of the precipitation–NDVI system also indicates that 1) long-term interactions (i.e., interannual and longer time scales) between the two anomalies tend to enhance one another, 2) short-term interactions (less than 2 months) tend to damp one another, and 3) intermediary-period interactions (4–8 months) are oscillatory. Together, these results support the hypothesis that vegetation may influence summertime climate variability via the land–atmosphere hydrological cycles over these semiarid grasslands.