A student collects soil moisture data from an instrument in the field.
A graduate student at Iowa State University collects data from a soil moisture instrument in an agricultural field. Field-based measurements are more accurate than satellite measurements but are impractical for deriving global soil moisture estimates. Credit: Lori Abendroth, Iowa State University
Source: Water Resources Research

Many modern insights into the global climate system stem from climate models, which use mathematical equations to represent physical processes in Earth’s atmosphere, oceans, land, and ice. Global climate models evaluate the energy balance between these four climate components to generate predictions and simulations of past and future conditions.

Land surface models are one component of a climate model. These models represent energy and mass exchanges between the land and atmosphere and portray many interactions, for example, the swap of carbon dioxide between plants and the atmosphere.

Soil moisture turns out to be a key variable in land surface models. It plays a pivotal role in water and biogeochemical cycles and helps to toggle the exchange of energy, or latent heat flux, between the surface of Earth and the lower atmosphere. However, the degree to which soil moisture controls latent heat is debatable, and the relationship remains difficult to quantify in climate models.

In a new study, Lei et al. tackle the relationship between soil moisture and latent heat flux. Using a novel, sophisticated statistical technique known as triple collocation, the authors offer an unbiased estimate of the coupling between soil moisture and latent heat. It is the first study to provide a robust, global assessment of this linkage.

Climate models use observations to develop representations of soil moisture and latent heat coupling. However, these observations always contain random errors, particularly when they are derived from remote sensing. This error results in biased estimates of soil moisture and latent heat coupling. The triple-collocation method implemented in the study works around this problem by sourcing data from three independent sources—referred to as triplets—that allowed the authors to quantify the variance of the errors and compensate for them when sampling estimates of coupling strength. To be included in the triple-collocation analysis, the data must derive from at least three independent sources. Several American and European Earth-observing missions supplied the data in this study.

In quantifying new estimates of soil moisture–latent heat flux coupling, the authors draw several significant conclusions. First, the study reveals that estimates of coupling strength derived from remote sensing–based estimates badly underestimate the true coupling strength. In contrast, estimates from land surface models overpredict the strength of the relationship.

Furthermore, the analysis identifies regional variations in these biases. For instance, the soil moisture–latent heat relationship is overpredicted by models in transitional areas between wet and dry climates, like those found in India, coastal Australia, and North America’s Great Plains.

The methods presented in the study could have far-reaching implications for climate models and predictions. The approach could help redefine how scientists represent a vital connection between the land and atmosphere in climate models. (Water Resources Research, https://doi.org/10.1029/2018WR023469, 2018)

—Aaron Sidder, Freelance Writer


Sidder, A. (2019), Are soil moisture and latent heat overcoupled in land models?, Eos, 100, https://doi.org/10.1029/2019EO119259. Published on 09 April 2019.

Text © 2019. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.