Meet our researchers! Tristan Quaife is Professor of Earth System Science at the University of Reading. His expertise lies in modelling of the terrestrial carbon cycle, radiative transfer in vegetation and data assimilation for land surface and ecosystem models. The University of Reading leads Task 1.2 in WP1, which focuses on understanding how the land surface responds to climate extremes. The EXPECT project is contributing by enhancing the integration of Earth Observation (EO) data to enable more accurate soil moisture modelling.
Soil moisture: a key variable in monitoring the impacts of climate change
“Part of the work that we do is to tune the soil hydraulics in a land surface model so that it provides a better match to the satellite data,” says Tristan Quaife, Professor of Earth System Science at the University of Reading.
Soil moisture plays a crucial role in a well-functioning climate warning system because it regulates the exchange of water and heat between land and atmosphere. It also determines how precipitation is divided – either soaking into the ground and running off the surface – affecting weather patterns and river flows.
Soil moisture is also important for the carbon cycle because the amount of water in soils directly influences vegetation’s ability to photosynthesize and take up carbon from the atmosphere.
How is EXPECT trying to achieve that?
All weather and climate models include some representation of the land surface beneath them: this is typically referred to as a land-surface model. These models control the fluxes of mass and energy between the land surface and the atmosphere, notably heat and water.
A key part of that is being able to model the hydraulics of the soil, namely how well water can enter, move through and be retained by soil. In EXPECT, we will adjust the hydraulic parameters of a land surface model to improve its alignment with satellite-based soil moisture observations.
One major challenge is that weather and climate models rely on accurate information about soil hydraulic parameters to make predictions. The lack of this information on a large scale limits the scope of weather forecasts, climate projections and early warning systems capabilities.
What is SMAP?
“At the moment, we’re working with data from SMAP, which is a NASA satellite, but as part of the EXPECT project, we are intending to move over to using the European Space Agency (ESA)’s Climate Change Initiative Soil Moisture Product,” says Prof. Quaife.
The ESA’s Climate Change Initiative (CCI) provides a range of climate-related data products, and one of these focuses specifically on soil moisture. This transition will allow us to leverage European data resources to improve our research.
A use case of land-model and EO data integration
“We know that soil moisture plays an important role in feeding back on local climate. So – for example – where we’ve got very dry soil, we’re more likely to get heatwaves and where we’ve got wet soil, less so. So, being able to tune our land surface models to represent soil moisture better should provide better predictions of climatic extremes,” says Prof. Quaife.
In the presence of very dry soils, the likelihood of experiencing heatwaves is considerably higher. This is especially concerning in southern Europe, where the scorching heatwaves frequently lead to crop failures and, in turn, to serious social and economic consequences.
Some communities in Africa already are beneficiaries of short-term agricultural forecasts based on a previous version of the soil moisture assimilation product. Similarly, the EXPECT research could deliver timely agricultural insights in Europe, which may support decision-making for farmers or agricultural planning in near real-time.
“We know that soil moisture plays an important role in feeding back on local climate. So – for example – where we’ve got very dry soil, we’re more likely to get heatwaves and where we’ve got wet soil, less so. So, being able to tune our land surface models to represent soil moisture better should provide better predictions of climatic extremes.”
How do we aim to interface models and data in EXPECT?
Models are never perfect, and observations come with their own uncertainties. The goal is to combine these imperfect models and observations to make the most accurate estimates of reality. Our models are large, complex, and computationally demanding, requiring advanced mathematics to integrate them effectively with observations.