Predictability study of heavy rainfall events in the Serra do Mar

Abstract

The Serra do Mar is a region of strategic importance to the State of São Paulo, both, for the sustainable development, due the remaining portions of the Mata Atlântica (the native forest), and for the economic development due to the railways, highways, fuel pipelines and the industrial and parI installations. However, this region suffer from frequent landslides which cause several material and life losses. These events have hydrometeorological causes in addition to the strong slope of the hills and the antropic actions. This project proposes the development of a monitoring and prediction system of risks in the Serra do Mar. The project is composed of 6 subprojects entitled: (1) "High resolution atmospheric modelling of extreme events in the Serra do Mar", (2) "Coupling an atmospheric model to a hydrological model", (3) "Characteristics of convective systems that cause extreme events in the Serra do Mar", (4) "Large scale characteristics associated with extreme events in the Serra do Mar", (5) "Development of a semi-automatic prediction and hydrometeorological information system in support of environmental disaster and risk managements", and (6) "the impact of geotechnical and hydrometeorological information from automatic surface stations on the forecasts over Serra do Mar region". The large scale atmospheric environment which characterizes a risky situation will be determined from calculations of atmospheric parameters and cluster analysis. The monitoring of the development, the trajectory and temporal evolution of the deep convective systems will be based on satellite imagery techniques for deep cloud morphology detection. This technique will provide Nowcasting of these weather systems. Wind forecasts from the mesoscale atmospheric model will be tested as input to this remote detection system in order to extend the forecast time range. Selected cases of extreme events will be simulated by the high resolution atmospheric and distributed hydrological models, Eta and Topog, respectively. Higher vertical and horizontal resolutions will be tested, the atmospheric non-hydrostatic effects will be evaluated. the sensitivity of the model to different convective and microphysics schemes and to land cover/use will be experimented to achieve better simulations. these numerical experimentations will be carried out in order to obtain an adjusted configuration of the models to produce 72 h forecasts. The Topog model will using observed and predicted information from the Eta modeI to provide river discharge forecast and indicate risks for landslides or floodings. The Topog will be coupled to the Eta model aiming at improving the rain and river discharge forecasts. Ensemble forecasts will be carried out to provide the information of probability of event occurrence. The proposed meteorological and hydrological mesonet- will provide real time information to capture the local circulation, heavy rains, the level of criticaI rivers and indication of regions of landslide risks. The observations are necessary not only for monitoring the rains, but also for better adjustment of the models and increase our understanding of the related phenomena. A database with socio-economic and geomorpholocial information will be implemented and combined with real time observations and numerical forecasts into a georreferenced information system. This database will be made available through the internet, so that the regions and the types of risk can be imediately identified. The outcome of this project is a prototype of monitoring and forecasting system developed for the risks due to heavy rains in mountainous regions. (AU)