Climate change is expected to increase the frequency and intensity of extreme rainfall events, which can have significant impacts on society and the environment.
However, current climate models are not very accurate in simulating these events, especially in the tropics and subtropics.
Scientists are trying to improve the representation of the physical processes that govern the formation and evolution of clouds and precipitation in these models, which could lead to more reliable projections of future changes in rainfall extremes.
The Challenge of Simulating Precipitation Extremes
PASCAL POCHARD-CASABIANCA/AFP via Getty Images
Precipitation extremes are defined as very high or low amounts of rainfall that occur over a short period of time, such as a day or an hour.
These events are influenced by many factors, such as the amount of water vapor in the atmosphere, the temperature and pressure gradients, the wind patterns, the topography, and the interactions between clouds and the environment.
Precipitation extremes are also highly variable in space and time, which makes them difficult to observe and measure.
Climate models are mathematical tools that simulate the behavior of the atmosphere, the ocean, the land, and the ice on a global scale.
They are based on the physical laws that govern the movement and exchange of energy and matter among these components.
These models are useful for understanding the past and present climate, as well as for projecting the future climate under different scenarios of greenhouse gas emissions and other human influences.
However, climate models have limitations in resolving the small-scale processes that are important for precipitation extremes, such as the formation and breakup of cloud droplets and ice crystals, the vertical motions of air parcels, and the microphysical interactions between clouds and precipitation.
These processes occur at scales of meters to kilometers, which are much smaller than the typical grid size of climate models, which ranges from tens to hundreds of kilometers.
Therefore, climate models have to use simplified representations of these processes, called parameterizations, which are based on empirical or theoretical relationships.
The choice of parameterization can have a large impact on the simulated precipitation extremes, as different parameterizations can produce different results for the same input data.
For example, some parameterizations may produce more intense rainfall events than others, or may shift the location or timing of the events.
Moreover, some parameterizations may not capture the feedbacks and interactions between the small-scale processes and the larger-scale circulation, which can also affect the precipitation extremes.
One way to improve the simulation of precipitation extremes in climate models is to use a technique called superparameterization, which involves embedding a high-resolution model of clouds and precipitation within each grid column of the climate model.
This allows the climate model to resolve the small-scale processes more explicitly, without relying on parameterizations.
Superparameterization can also capture the feedbacks and interactions between the clouds and precipitation and the larger-scale circulation, which can enhance the realism of the simulated precipitation extremes.
It has been shown to improve the simulation of precipitation extremes in various regions of the world, such as the tropics, the extratropics, and the monsoon regions.
Superparameterized models can reproduce the observed patterns and statistics of precipitation extremes more accurately than conventional models, as well as the physical mechanisms and factors that drive these events.
These can also simulate the changes in precipitation extremes under global warming scenarios more consistently and robustly than conventional models
However, superparameterization is not a perfect solution, as it still has some limitations and uncertainties.
For example, superparameterization is computationally expensive, as it requires running many high-resolution models within the climate model, which increases the computational time and resources.
It also depends on the choice of the high-resolution model and its configuration, such as the grid size, the time step, and the boundary conditions, which can affect the simulated precipitation extremes.
Moreover, superparameterization does not resolve all the small-scale processes that are relevant for precipitation extremes, such as the subgrid-scale variability and turbulence, which may still require parameterizations.
Therefore, superparameterization is not a substitute for parameterization, but rather a complement that can help improve the representation of the physical processes that govern precipitation extremes in climate models.
Superparameterization can also provide insights and guidance for developing and evaluating new parameterizations that can better capture the behavior and variability of precipitation extremes.
It can also help identify the sources and uncertainties of the simulated precipitation extremes, which can inform the interpretation and communication of the model results and projections.
