Features:

BRAZIL IS IN NEED OF "AMAZONIA SIMULATOR"
By: Christopher Lazou, HiPerCom Consultants, Ltd.

The Brazilian state-owned institute for space research, INPE (Instituto Nacional de Pesquisas Espaciais), is installing 12 Nodes (96 processors) NEC SX-6. The peak performance of the total system is 768 Gigaflop/s, the most powerful system sold by NEC in South America. The final phase of the installation is to be completed last quarter 2003. The SX-6 is to be used for weather forecasting and climate research by CPTEC (Centro de Previsão de Tempo e Estudos Climáticos). The new technology and power of the NEC SX-6 is expected to substantially improve the accuracy of Brazilian weather and climate forecasts.

In my recent visit to Brazil, I had the opportunity to interview Dr. Carlos A. Nobre, Director of CPTEC and scientific co-ordinator of the Large Scale Biosphere-Atmosphere Experiment in Amazonia, (LBA). I feel it is of sufficient interest to the HPC community to merit an article. Below is a synopsis of this interview with Dr. Carlos A. Nobre, Director of Brazilian Weather Forecasting and Climate Studies Centre, Cachoeira, Paulista, Brazil, CPTEC, contacted on 21 May 2003.

Lazou: When was CPTEC set up and who funds it?

Nobre: CPTEC was set up in 1994, as part of INPE, the Brazilian state owned institute of space research and is funded by the Brazilian government.

Lazou: What is CPTEC's Mission?

Nobre: CPTEC's mission is to provide Brazil with state-of-the-art weather forecasts and climate predictions for the benefit of society. The initial goal was to modernise weather forecasts in Brazil. Planning for such modernisation started in 1988. The installation of the NEC SX-3 in 1994 enabled us to produce daily weather forecasts using the atmospheric General Circulation Model (GCM). This was followed by seasonal and regional models and by 1998, with the installation of the SX-4, we were able to utilise ensemble seasonal climate forecasts as well as increased model resolution. By 2001, we coupled the atmosphere, ocean, hydrology, and wave models. Now with the installation of the SX-6, we are using AGCM for 50 years climate simulations, using higher resolution ensembles and generally improving our weather forecasts and climate simulations. Further details can be found on CPTEC's Web home page: http://www.cptec.inpe.br.

Lazou: With the Amazonian rainforest and the Andes, the largest mountain range in the world at the other end, Brazil must have a very challenging weather and climate. Can you give a brief description of any projects you are involved. In particular, the Large Scale Biosphere-Atmosphere Experiment in Amazonia, (LBA).

Nobre: Amazonia is the last great tropical forest domain and the largest hydrological system on the planet. Highly integrated, these natural systems play an important role in the function of regional and global climate. The LBA programme integrates knowledge to understand the Amazon basin, how do these regional environmental systems function and how do changes in natural Amazonian cycles affect life on Earth. For example, how does the Amazon terrestrial biosphere interact with the atmosphere? Is the Amazon forest a source or a sink of carbon? What is the re-growth capacity of deforested areas?

Impacts on the ecosystems may produce irreversible effects and change in the climate of the region, the continent and the world. The understanding of the Amazonian ecosystems will contribute to the development of techniques for sustainable use of the land in the Amazon environment. Scientific research, including field data collection, is still necessary to find answers to the many questions about regional and global impacts of human activities.

Lazou: What are the fundamental objectives of LBA?

Nobre: To improve the understanding of the Amazonian ecosystems through complementary and integrated scientific studies in the following areas: Physical climate, hydrology and water chemistry, carbon exchange, nutrient cycling, atmospheric chemistry and trace-gases, land cover and land use change and the human dimension of all these. It encourages the participation of new researchers through grants for training and education, particularly in the Amazon region and makes research results available for the support and promotion of the development of alternative conservation and sustainable land- use practices in Amazonia.

Lazou: We talk about Amazonia, but what is the size and biosphere content of this region?

Nobre: The natural system of Amazonia consists of almost 6 million square kilometres of contiguous tropical forest, containing perhaps a third of the planets biodiversity. It has abundant rainfall, 2.2 metres annually, providing 18% of freshwater input into the global oceans (220,000 cubic metres per second). Over 120 Gigatons of carbon is stored in vegetation and soil, has a multitude of ecosystems, biological and a strong ethnic diversity.

Lazou: What is the recent history of Amazonia and why should we be interested in it now?

Nobre: Land use change in Amazonia from year 1500 up to the 1960's was negligible, but it sustained large loss of ethnic diversity due to colonisation. The large land use change, occurred in the last 30 years. Close to 600,000 square kilometres, 15% of Brazilian Amazonia is deforested. From 15,000 up to 30,000 square kilometres per year of forest is burned to convert land into pasture for beef production. This has the additional hazard of many chemicals aerosols being released into the atmosphere.

The government plans to "integrate" Amazonia with a road network throughout the region. The population has grown from 3.5 million to 20 million since 1970 though 65% live in large and mid-size towns. The colonization projects of landless people using small-scale low technology agriculture, subsidized cattle ranching and destructive logging, are some of the vectors to subsequent deforestation.

Lazou: When did the LBA project start, how is funded and who participates?

Nobre: The plans for LBA were in place in 1994, when the first supercomputer, the NEC SX-3 was purchased for weather forecasting in Brazil, but the experiment started in 1998. The LBA project gets its funds from the Brazilian government, NASA in the USA and from eight European countries through the EU- Commission and national contributions. The total funds are 80 million US dollars, over a period of six years. LBA is an important component of the international studies on global change. Some 500 researchers are taking part from many countries.

Lazou: How is, the LBA experiment performed, how is the data collected and who looks after the data gathered?

Nobre: Atmospheric and ground data are obtained from the entire region using towers equipped with instrumentation, which measure greenhouse-gas fluxes, oxidants and aerosols. These are complemented by actual tree growths using sampling statistical techniques. Satellite imagery is also used and these methods create a large amount of data. The LBA-DIS (Data and Information System), operated in Brazil, with a mirror site in the USA, provides storage, a search-engine, and the sharing of data gathered. Data collection should be completed by year 2005.

Lazou: What happens next and how are you going to analyse these data?

Nobre: The supercomputer at CPTEC is a very important resource, for synthesis of several models. To couple physical, bio-geophysical and geo-biochemical to the global model, needs a very large computer. We are in need of an "Amazon Simulator" by year 2010, similar to the Earth Simulator presently installed in Japan.

Lazou: Are there any preliminary results you can share. Is Amazonia a carbon source or a sink?

Nobre: Over 120 Gigatons of carbon is stored in the Amazonian vegetation and soil. Emissions, resulting from deforestation, are around 200 Megatons per year. The balance of evidence provided by an array of different and independent observations (inverse calculations, forest inventories, carbon pools and flux studies, flux towers measurements, soil chambers, boundary layer budgeting, land use change, fire emissions and so on) still does not allow for firm conclusions, but currently it cannot be ruled out the possibility that Amazonia functions as a sink of carbon. This is estimated to be between 0.3 to 0.5 Gigatons per year. The estimate for total terrestrial biota sink is about 3.0 Gigatons, but saturation is expected by year 2050 due to global warming.

Another simulation study is using GCM modelled deforestation and climate interactions. Can vegetation type, i.e. pasture or forest, determine or modify regional or global climate? Large scale deforestation according to the numerical simulation would result to 1 to 2.5 Centigrade increase in surface temperature, 15% to 30% decrease in evaporation, 5% to 20% decrease in rainfall, concentrated mostly in the dry season, having a tendency to "Savanization" of Amazonia.

Also a study by Professor Brian Hoskins, University of Reading, UK, has shown that changes in Amazonia, affects the weather in the UK and Scandinavia. Dry conditions in Amazonia in September/October, causes a wet UK and a dry Scandinavia. These observed and modelled tele-connections need to be explored further and for this a large-scale supercomputer is needed.

Lazou: Finally, people are concerned about climate change causing "desertification" with catastrophic effects on South America and the rest of the world. Are these LBA studies, shedding any light on this?

Nobre: Natural ecosystems in tropical Brazil have been under increasing land use change pressure (e.g. Amazonia deforestation and Northeast Brazil "desertification"). These land use changes, would cause reduction of rainfall in both Amazonia and Northeast Brazil. The synergetic combination of regional climate changes, caused by global warming and by land use change over the next several decades, could tip the biome-climate state to a new stable equilibrium with "Savanization" of parts of Amazonia and "desertification" in Northeast Brazil.

Lazou: Thank you for sharing your experience about the Brazilian climate research. I know that Brazil is struggling to feed its people and saddled with some 40 billion US dollars bill just to service its debt. This makes it imperative that your Amazonia findings are of sufficient quality to help your government to assess the impact of present actions and investments, focus its priorities to meaningful programmes for societal development, bringing social and economic benefits to the Brazilian people. I hope funds from developed countries, Europe, and the USA, are forthcoming to install an "Amazon Simulator", so you can continue climate studies for the benefit of the world.

(Brands and names are the property of their respective owners) Copyright: Christopher Lazou, HiPerCom Consultants, Ltd., UK. June 2003.