|Abstracts on Global Climate Change|
Periodic solutions for soil carbon dynamics equilibriums with time-varying forcing variables
Martin, MP Cordier, S Balesdent, J Arrouays, D
ECOLOGICAL MODELLING 204:3-4 523-530
Numerical models that simulate the dynamics of carbon in soil are increasingly used to improve our knowledge and help our management of the carbon cycle. Calculation of the long-term behavior of these models is necessary in many applications but encounters the difficulty of managing the periodic forcing variables, e.g. seasonal variations, such as carbon inputs and decomposition rates. This calculation is conventionally done by running the model over large time durations or by assuming constant forcing variables. Two methods, which make it possible to rapidly compute periodic solutions taking into account the time variations of these variables, are proposed. The first one works on discrete-time models and the second one on continuous-time models involving Fourier transforms. Both methods were tested on the Rothamsted carbon model (RothC), a discrete-time model which has also been given a continuous approximation, using realistic and unrealistic sets of time-varying forcing functions. Both methods provided an efficient way to compute the periodic solutions of the RothC model within the application domain of the model. Compared to running the discrete model to the equilibrium, reduction in the computational cost was of up to 95% at the expense of a maximum absolute error of 1% for the estimation of carbon stocks. For specific distributions of the forcing variables the use of Fourier transform of zero order, which was equivalent to assume constant forcing variables, led to a maximum absolute error of SS% in the estimation of the long-term behavior of the model. There, a Fourier transform of order higher than zero is required. (C) 2007 Elsevier B.V. All rights reserved.
Simulation of seasonal precipitation and raindays over Greece: a statistical downscaling technique based on artificial neural networks (ANNs)
Tolika, K Maheras, P Vafiadis, M Flocasc, HA Arseni-Papadimitriou, A
INTERNATIONAL JOURNAL OF CLIMATOLOGY 27:7 861-881
A statistical downscaling technique based on artificial neural network (ANN) was employed for the estimation of local changes on seasonal (winter, spring) precipitation and raindays for selected stations over Greece. Empirical transfer functions were derived between large-scale predictors from the NCEP/NCAR reanalysis and local rainfall parameters. Two sets of predictors were used: (1) the circulation-based 500 hPa and (2) its combination along with surface specific humidity and raw precipitation data (nonconventional predictor). The simulated time series were evaluated against observational data and the downscaling model was found efficient in generating winter and spring precipitation and raindays. The temporal evolution of the estimated variables was well captured, for both seasons. Generally, the use of the nonconventional predictors are attributed to the improvement of the simulated results. Subsequently, the present day and future changes on precipitation conditions were examined using large-scale data from the atmospheric general circulation model HadAM3P to the statistical model. The downscaled climate change signal for both precipitation and raindays, partly for winter and especially for spring, is similar to the signal from the HadAM3P direct output: a decrease of the parameters is predicted over the study area. However, the amplitude of the changes was different. Copyright (c) 2006 Royal Meteorological Society
A maximum entropy method for combining AOGCMs for regional intra-year climate change assessment
Laurent, R Cai, XM
CLIMATIC CHANGE 82:3-4 411-435
This paper deals with different responses from various Atmosphere-Ocean Global Climate Models (AOGCMs) at the regional scale. What can be the best use of AOGCMs for assessing the climate change in a particular region? The question is complicated by the consideration of intra-year month-to-month variability of a particular climate variable such as precipitation or temperature in a specific region. A maximum entropy method (MEM), which combines limited information with empirical perspectives, is applied to assessing the probability-weighted multimodel ensemble average of a climate variable at the region scale. The method is compared to and coupled with other two methods: the root mean square error minimization method and the simple multimodel ensemble average method. A mechanism is developed to handle a comprehensive range of model uncertainties and to identify the best combination of AOGCMs based on a balance of two rules: depending equally on all models versus giving higher priority to models more strongly verified by the historical observation. As a case study, the method is applied to a central US region to compute the probability-based average changes in monthly precipitation and temperature projected for 2055, based on outputs from a set of AOGCMs. Using the AOGCM data prepared by international climate change study groups and local climate observation data, one can apply the MEM to precipitation or temperature for a particular region to generate an annual cycle, which includes the effects from both global climate change and local intra-year climate variability.