Study Will Make Storm Forecasting More Accurate
It follows an article published on day (07/22) in the english website of the Agência FAPESP noting that study will make Storm Forecasting more accurate.
Study Will Make Storm
Forecasting More Accurate
By Karina Toledo
July 22, 2015
(Photo: Wikimedia Commons)
Real and simulated data were compared using an innovative
tracking technique that computed the sizes and lifetimes of cloud
and rain cells, comparing simulated clouds with satellite and radar data.
– In a study by , a team of French and Brazilian researchers identified and corrected a flaw in the mathematical models used to simulate cloud and rain formation processes.
According to the authors, the study will make storm forecasting more precise. “We compared a simulation produced by a high-resolution model with observational data collected in 2012 in Santa Maria, a town in Rio Grande do Sul, southern Brazil. The region has some of the most severe storms on earth,” said Luiz Augusto Toledo Machado, a researcher at Brazil’s National Space Research Institute (INPE).
“We noticed that many small clouds appearing in the model weren’t in fact observed using satellites and rain radars,” Machado added. “So we decided to investigate why this was so.” The study was part of a supported by FAPESP and coordinated by Machado.
The Santa Maria data collection operation was carried out for a major scientific campaign held from 2010-14 as part of (meaning “rain”). Its aim was to understand the physical processes that take place inside clouds, including variations in droplet size and other features of rain and cloud cells such as the proportions of water and ice, and the development of lightning and thunderstorms. All of this research is considered important to improve extreme event forecasting (read more at ).
The storms that typically form in this part of southern Brazil, Machado explained, are classed as mesoscale convective complexes (MCCs), and they were simulated using a French model called Meso-NH, short for non-hydrostatic mesoscale atmospheric model.
The study described in MWR was a collaboration between Machado and Jean-Pierre Chaboureau, a physicist at the CNRS Aerology Laboratory in Toulouse, France.
“The regional models that simulate cloud formation typically work with a scale of about 10 km, meaning they can generate data for points 10 km apart. Meso-NH generates data every 2 km, which is high-resolution in the sense that clouds can be simulated more precisely. This is the way to go in future. It will let us forecast rain in particular neighborhoods of a city, for example,” Machado said.
Real and simulated data were compared using an innovative tracking technique that computed the size and lifetime of cloud and rain distributions. Histograms enabled the researchers to compare the dimensions of simulated clouds with those of clouds observed via satellite and radar.
When they investigated mismatches between the simulated and real data, the researchers discovered that the model did not accurately represent entrainment, the mixture of environmental air into air currents and clouds.
“Entrainment occurs when a turbulent flow captures a non-turbulent flow. In the model, turbulence was parameterized in one dimension. We created a 3D parameterization and slightly increased the mixing length, which is the average distance an inbound air parcel travels before interacting with an air parcel inside the cloud,” Machado said.
The modifications made the distribution of simulated and actual cloud sizes and heights more similar. “This will undoubtedly have an impact on the quality of rainfall forecasting. We showed in a case study that accuracy increases when turbulence is corrected,” Machado said.
Previous studies had suggested that similar problems with other mathematical models of cloud formation could be remedied using this approach.
The article was one of the first produced by Project CHUVA, which included data collection campaigns in the cities of Alcântara (Maranhão), Fortaleza (Ceará), Belém (Pará), São José dos Campos (São Paulo) and Manaus (Amazonas), besides Santa Maria (Rio Grande do Sul). The regions chosen for the field research represent the different precipitation regimes found in Brazil.
Source: English Website of the Agência FAPESP