Metode pengujian model-model sirkulasi umum untuk wilayah Indonesia

Increasing CO₂ concentration in the atmosphere will cause change of climate. Studies on the impact of climate change have been conducted by several research agencies. This study discussed the description of five GCM models: CCCM, GISS, GFDL, UKMO and CSIRO9, and methodologies for evaluating the performance of the models in Indonesia. There are two methods being evaluated for the evaluation the performance of the five models. First, methods using correlation analysis and RMSIE (Root means square interpolation error). Second, using regression analysis with dummy variables. In the former method the evaluation based on montly data while the later method based on monthly patern of the data. Since the second approach considers not only the quantity of the data but also the pattern of the data, thus the second approach is suggested. However, the second approach required several steps of analysis while the first approach only required one step. Using the first approach, the better models were CSIRO9 and CCCM, while using second approach are CSIRO9 and GFDL. This result suggested that among the five models, CSIRO9 appears to be best GCM model for Indonesia. However, in general the five GCM models did not perform satisfactorily, since the level of similarity between observed and GCM-generated data are low. Therefore, further exploration to other GCM models is required.