Author(s): LUIZ, A. J. B.; MAIA, A. de H. N.

Summary: Resumo: Para a utilização segura das projeções de precipitação oriundas de modelos climáticos regionalizados no intuito de estudar os impactos de possíveis mudanças climáticas sobre recursos hídricos, em escala de bacia hidrográfica, é preciso que se tenha confiança que os modelos são capazes de reproduzir com consistência os dados reais. É aceitável que um modelo não reproduza exatamente a realidade e até apresente vieses passíveis de correção. Entretanto, especialmente na escala de bacias e sub-bacias brasileiras, carecemos de estudos que avaliem o desempenho desses modelos. Neste trabalho, avaliamos a consistência do modelo regionalizado Eta-HadGEM2-ES, com relação às projeções médias mensais de chuva (1999 a 2013), para as bacias dos rios Paracatu e São Marcos, na região central do Brasil, comparados aos dados interpolados fornecidos pelo Climate Research Unit (CRU) e dados reais obtidos de estações meteorológicas locais. As diferenças observadas no tempo e no espaço foram elevadas, tanto em valor absoluto quanto relativo e não apresentaram um padrão sistemático de desvios que permitisse a correção de viés. Abstract: For the safe use of precipitation projections arising from regional climate models in order to study the impacts of possible climate change on water resources in river basin scale it is necessary to have confidence that the models are able to reproduce consistently the actual data. It is acceptable that a model does not exactly reproduce the reality and to present biases that we can correct. However, especially on the scale of Brazilian?s basins and sub-basins, there are few studies to evaluate the performance of these models. In this paper, we evaluate the consistency of regional model Eta-HadGEM2-ES, with respect to the average monthly projections of rainfall (1999-2013), for the Paracatu and São Marcos river basins, in central Brazil, compared to the interpolated data provided by Climate Research Unit (CRU) and real data from local weather stations. The differences in time and space were high, both in absolute and relative value and did not show a systematic pattern of deviations to allow the bias correction. Abstract: For the safe use of precipitation projections arising from regional climate models in order to study the impacts of possible climate change on water resources in river basin scale it is necessary to have confi dence that the models are able to reproduce consistently the actual data. It is acceptable that a model does not exactly reproduce the reality and to present biases that we can correct. However, especially on the scale of Brazilian?s basins and sub-basins, there are few studies to evaluate the performance of these models. In this paper, we evaluate the consistency of regional model Eta-HadGEM2-ES, with respect to the average monthly projections of rainfall (1999-2013), for the Paracatu and São Marcos river basins, in central Brazil, compared to the interpolated data provided by Climate Research Unit (CRU) and real data from local weather stations. The diff erences in time and space were high, both in absolute and relative value and did not show a systematic pattern of deviations to allow the bias correction.

Publication year: 2016

Types of publication: Paper in annals and proceedings

Unit: Embrapa Environment

Keywords: Bacia Hidrográfica, Chuva, Clima, Climate models, Correção de Viés, Estudos de impacto, Previsão do tempo, Rain, Redução de escala, Weather forecasting