Due to global warming, the existing IDF curves that were derived from historical data may no longer valid for estimating future rainfall. On the other hand, the alternative future rainfall data generated from Community Climate System Model version 3 (CCSM3) cannot be used directly due to systematic biases caused by non-inclusion of local features. Quantile mapping was used to bias correct the rainfall data generated by CCSM3 model based on the properties of observed rainfall. Only 4 out of 16 stations in the state of Kelantan have shown an increasing significant trend for 1-h annual maximum rainfall. The best transfer function used to correct bias in CCSM3 data is the second order polynomial function. The quantile mapping for bias correction is satisfactory as the highest RMSE was only 4.045, a reduction by 15% compared to the original CCSM3 data. Subsequently, a unique transfer function was developed to represent the behavior of CCSM3 and observed data for each station. The future intensity was found to increase for short and medium return periods of 2-, 5-, 10- and 25-year, where most of the stations have Climate Change Factor (CCF) larger than 1. On the contrary, for a longer return period of 50- and 100-year, the intensities are predicted to be lower.

由于全球变暖，从历史数据中得出的现有IDF曲线可能不再对估计未来的降雨有效。另一方面，由社区气候系统模型版本3（CCSM3）生成的替代性未来降雨量数据由于未包含局部特征而导致的系统偏差而无法直接使用。基于观测到的降雨的性质，使用分位数映射对CCSM3模型生成的降雨数据进行偏差校正。吉兰丹州16个站中只有4个站的年最大降雨量1小时显示出显着增加的趋势。用于校正CCSM3数据偏差的最佳传递函数是二阶多项式函数。偏差校正的分位数映射令人满意，因为最高RMSE仅为4.045，与原始CCSM3数据相比降低了15％。随后，开发了独特的传递函数来表示CCSM3的行为以及每个站点的观测数据。人们发现，在2年，5年，10年和25年的短期和中期回归期，未来的强度会增加，其中大多数台站的气候变化因子（CCF）大于1。相反，更长的时间50年和100年的回归期，强度预计会降低。