ADVANCED COMPUTING IN ANALYSIS OF THE CLIMATE CHANGES IMPACT
This project is aimed at regional climate computer simulations for Bulgaria, thus contributing to the development and execution of national action plan for adaptation to climate changes. The global warming is a consequence of ongoing changes of the circulation of the system ocean-atmosphere, which lead not only to temperature changes, but also to the spatial/temporal distribution of precipitation, hence the global water budgets, to changes of the characteristics and spatial/temporal distribution of unfavorable and catastrophic events (drought, storms, hail, floods, fires, sea waves, soil erosion, etc.). The changes will have their influence on ecosystems, on all sectors of the economy and every aspect of human activity and quality of life. The regional/local characteristics of the climate changes cannot be predicted correctly by the global models. The action plan should be formulated for a given country on the basis of a prediction of the regional/local scale specifics in climate changes and their consequences. The objectives of the planned research is to develop adequate methodology and implementing it to produce reliable, comprehensive and detailed evaluations of possible regional/local climate changes and their consequences for different global change scenarios, thus providing scientifically robust basis for elaboration, permanent efficiency evaluation and upgrade of the action plan.. The main scientific challenges are the multi-scale nature of the processes and the complex interactions of different scale phenomena and basic mechanisms and pathways trough which regional/local climate change specifics and their impact on environment and human activities are formed. From the other hand detecting the various possible consequences of regional/local climate changes and evaluating them in unified metrics is a novel multidisciplinary task. Facing these challenges leads to extremely large-scale simulations, which requires developing and applying of advanced computing including Big Data, and High Performance Grid and Cloud technologies.
1. Creation of unified geo-information environment
The studies will be carried out applying the most comprehensive available Big Data. This is a huge amount of information, which can be utilized only by development over-arcing meta-data base with corresponding procedures for data access and processing. The studies should be carried out using up-to-date models (e.g. RegCM, WRF, ALADIN), well validated and tuned to the physiographic and climatic specifics of the Balkan Peninsula and Bulgaria by performing a 10 year hind cast for a comprehensive set of model configurations, which, compared with the available meteorological (climatic) data, to allow choosing the configuration, which best describes the regional/local specifics. The resources required for model tuning and validation are huge, so applying advanced high performance computing is the only way to accomplish the task.
2. Forecast of the future regional/local changes of the climate
The future climate changes in the country will be evaluated for time slices of 50 to 100 years ahead on the basis of a set of generally accepted global climate change scenarios. The spatial and temporal variability of the basic climatic parameters will be forecasted for the different global change scenarios. The spatial/temporal recurrence of extreme/catastrophic atmospheric events will be evaluated. Ensemble and multi-model simulations will be carried out, which again requires application of advanced computing technologies.
3. Evaluation of the climate change impact on ecosystems, economy and quality of life
The evaluated climate changes will have their impact on all the other components of the environment – water cycles, soils, ecosystems, biodiversity, air and water quality, etc. and hence on economy, quality of life and in general any human activity aspect. These impacts will be evaluated using up-to-date models for Big Data analytics in the terms of unified metrics. The proper choice of this set of models and introducing the universal metrics is the critical challenge of this task.