The Amazon rain forest accounts for 60% of the world’s remaining rain forests. The biodiverse region spans more than 5.5 million square kilometers (2.1 million square miles) and is often referred to as the planet’s lungs because of its capacity to suck in atmospheric carbon dioxide. In addition to its central role in the global carbon cycle, the Amazon shapes water and energy budgets around the world by transferring vast amounts of water from the land to the atmosphere through the process of evapotranspiration.
Evapotranspiration is a tricky process to observe, particularly over large and diverse landscapes and during times of water scarcity. Generally, scientists use models to simulate evapotranspiration across different scales. These models require extensive data inputs that describe both weather conditions and vegetative characteristics, and these variables may carry a significant degree of uncertainty. Over areas less than 1 square kilometer, ground-based monitoring stations such as FLUXNET are used to collect these data. For larger areas, satellites provide relevant data, for example, the Moderate Resolution Imaging Spectroradiometer (MODIS) on board the Terra and Aqua satellites.
Xu et al. have attempted to simplify models estimating evapotranspiration over the Amazon. The authors introduced a new approach based on the theory of maximum entropy production that uses only three inputs: net radiation, air humidity, and temperature. These data are more accessible than hard-to-measure variables such as stomatal resistance and vapor pressure deficit required by other models. This new method was previously used to estimate evapotranspiration but not in diverse environments like the Amazon.
The authors chose to use data from nine FLUXNET towers in the Amazon and compared the results to a conventional model that relies on remotely sensed data from MODIS. The results suggest that the maximum entropy production model adequately simulates evapotranspiration across the sites in the rain forest. The model provided consistently better estimates of evapotranspiration than other models and did so under different water stress situations.
Comparatively, the conventional model relying on MODIS data did not perform as well. The MODIS-based model tended to underestimate evapotranspiration at low rates while overestimating it in higher ranges.
The results suggest that the maximum entropy production approach can adequately estimate evapotranspiration in biodiverse tropical regions and even outperforms the evapotranspiration products available from MODIS that are commonly used by hydrologists. Furthermore, the simplicity of the model is an attractive alternative to conventional approaches that rely on hard-to-measure data with substantial uncertainty. (Geophysical Research Letters, https://doi.org/10.1029/2018GL080907, 2019)
—Aaron Sidder, Freelance Writer