Soil structure is an important omission in Earth System Models

Soil plays a crucial role in Earth's life, linking atmospheric and hydrologic processes with the biosphere and biogeochemical cycles. Approximately 40% of terrestrial precipitation returns to the atmosphere through the soil-plant-atmosphere continuum, and nearly all terrestrial annual global vegetation production relies on soil processes. Soil properties, vegetation attributes, and land-use patterns jointly influence surface energy and water fluxes, regional climate, and extreme events like heatwaves and droughts. Soil moisture and texture particularly affect weather patterns, surface evaporation, and temperature extremes. Soil hydraulic properties govern soil water fluxes, impacting groundwater and surface water resources. The challenge lies in representing soil processes at regional and global scales within Earth System Models (ESMs), particularly the effects of soil structure. Even at the profile scale, quantifying soil structure's influence on infiltration and water fluxes is difficult and often neglected. Current ESM parameterization of soil hydraulic properties heavily relies on easily measured soil textural maps, largely ignoring soil structure and pedogenic information. This omission is inherent in pedotransfer functions (PTFs), used to deduce soil hydraulic functions from primarily soil textural information. While the influence of soil structure on soil hydraulic functions and soil ecology is well-established, few studies have incorporated it into PTFs. This research aims to bridge this gap by assessing the significance of soil structure omission in hydro-climatic modeling.

The paper extensively reviews the existing literature on soil hydraulic properties and their representation in Earth System Models. It highlights the prevalent use of pedotransfer functions (PTFs) that primarily rely on easily measurable soil textural properties, neglecting the significant role of soil structure. The authors cite numerous studies demonstrating the impact of soil structure on various hydrological processes, including infiltration, runoff, and drainage. They discuss the limitations of existing PTFs in capturing the complex interactions between soil structure, biological activity, and hydrological responses. The review also points to the scarcity of studies that explicitly incorporate soil structure into PTFs, emphasizing the need for improved parameterization schemes in ESMs. The authors highlight the bias introduced by the limited and geographically clustered soil samples used to train PTFs, which primarily focus on agricultural soils, overlooking the diversity of natural landscapes.

To address the omission of soil structure in ESMs, the researchers developed a simple parameterization to incorporate soil structure of biophysical origin. They focused on the influence of soil structure on soil hydraulic conductivity and, to a lesser extent, water retention. These modifications were approximated and used to assess their impacts on subsurface and surface fluxes at local and global scales. The modified soil parameterization was tested in two ways. First, they used the Tethys-Chloris (T&C) ecosystem model to assess eco-hydrologic responses at 20 globally diverse locations. Hourly meteorological observations (3–31 years) were used as model inputs, with vegetation parameters representative of local conditions. Three soil scenarios were simulated: (1) soil hydraulic parameterization based on the SoilGrids-250m database and Tóth PTFs without soil structural effects; (2) the same global map but with soil structural effects; and (3) original soil hydraulic parameterization based on local soil textural properties and Saxton and Rawls PTFs. Runoff and drainage were analyzed. Second, the Ocean-Land-Atmosphere Model (OLAM) was used for global-scale assessments. A 35-year simulation was run with and without soil structure, preceded by a century-long spin-up period. Land-surface fluxes and climate variables were compared between the simulations. Statistical two-sample t-tests assessed significant differences for 11 climatic and hydrological variables across 27 geographical regions and 12 months. The ratio between structural and textural saturated conductivity (Ks,str/Ks,tex) was parameterized as a function of Gross Primary Production (GPP), used as a proxy for bioturbation activity. The impact of soil structure diminished with depth, proportionally to the cumulative root depth distribution.

At the local scale, including soil structure significantly modified infiltration-runoff partitioning and recharge in wet and vegetated regions. Increased drainage and reduced surface runoff were observed, particularly in productive ecosystems and poorly drained soils. For instance, at a tropical rainforest site, accounting for soil structure increased drainage by 1050 mm/year and decreased surface runoff by 1280 mm/year, representing roughly 40–45% of annual precipitation. Changes were also considerable in semi-arid ecosystems, representing a substantial percentage of the water balance. However, the vertical distribution of soil moisture was also significantly altered, with increased water content in deeper soil layers due to enhanced vertical redistribution in wet conditions. Surface soil moisture remained unaffected, suggesting that surface evaporation and root water uptake dampened the effects of soil structure. The effects on energy fluxes, transpiration, and vegetation metrics were relatively small (<2%), except in two sites where latent heat changes reached up to 15%. At the global scale, however, differences between simulations with and without soil structure were statistically insignificant for most variables. This was attributed to the coarse spatial resolution of global-scale analyses, which masks the effects of short-intense rainfall events and lateral water redistribution.

The study successfully demonstrated the importance of incorporating soil structure into hydrological models at the ecosystem scale, showing significant effects on runoff and drainage. However, the lack of significant global-scale effects highlights a critical limitation of current ESMs. The coarse spatial resolution of these models likely prevents the capture of fine-scale processes associated with soil structure, especially those triggered by intense rainfall events. While the study's findings demonstrate the importance of soil structure at local scales, their limited impact on global-scale climate simulations suggests the need for improvements in model resolution and the representation of hydrological processes, particularly lateral water fluxes. The study emphasizes the conditional nature of soil structure effects, which depend on local climatic and soil textural conditions.

This research provides compelling evidence for the significant impact of soil structure on local hydrological processes. The lack of detectable global-scale effects emphasizes the limitations of current ESMs' coarse resolutions. The study highlights the need for higher-resolution models to capture the effects of soil structure on climate dynamics and water resource management. Future research should focus on improving model resolution, incorporating more detailed representations of hydrological processes, and refining soil parameterization schemes by integrating additional data on soil structure and its relationships with bioturbation. Further investigation into the impact of soil structure on soil biogeochemistry is also warranted.

The study acknowledges several limitations. The parameterization of soil structure effects was simplified, using GPP as a proxy for bioturbation activity. This simplification might not capture the complexity of soil structure formation and its interactions with different soil textures. The coarse resolution of global-scale simulations may have masked the true impact of soil structure on global climate patterns. The lack of detailed information on soil structure at global scales hinders a more comprehensive assessment. Future studies should employ higher resolution models and incorporate more detailed data to overcome these limitations.