Premium
Ranking the AR4 climate models over the Murray‐Darling Basin using simulated maximum temperature, minimum temperature and precipitation
Author(s) -
Maxino C. C.,
McAvaney B. J.,
Pitman A. J.,
Perkins S. E.
Publication year - 2007
Publication title -
international journal of climatology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.58
H-Index - 166
eISSN - 1097-0088
pISSN - 0899-8418
DOI - 10.1002/joc.1612
Subject(s) - precipitation , climatology , climate change , environmental science , climate model , probability density function , ranking (information retrieval) , mean radiant temperature , structural basin , function (biology) , meteorology , statistics , econometrics , computer science , mathematics , geography , geology , paleontology , oceanography , machine learning , evolutionary biology , biology
We assess the capacity of models submitted for the Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4), to simulate the maximum and minimum temperatures and precipitation over the Murray‐Darling Basin (Australia). We use daily data from the AR4 to calculate the mean of these three quantities, but we also derive probability density functions (PDFs) for each variable. We quantify the skill of each climate model to simulate the probability density function as a basis of assessing those models with significant capacity over the basin. We show that Commonwealth Scientific and Industrial Research Organization (CSIRO), Insitut Pierre Simon Laplace (IPSL), and MIROC‐m capture the observed PDFs of maximum and minimum temperature and precipitation relatively well. Other models capture one or two of these variables well but show limitations, or could not be assessed, in a third. We, therefore, recommend these three models to users of model results, but note that this recommendation is limited to this basin. However, our methodology provides quite a straightforward and quantitatively based means to choose climate models for impact assessment in any data‐rich region. Specifically, we note that to demonstrate skill in simulating the daily derived PDFs is far more challenging than just simulating the mean (as a model can simulate the mean well by averaging large biases of opposite sign). Our method therefore provides a more robust foundation for using a model in impacts assessment. Copyright © 2007 Royal Meteorological Society