Complex topography and wind characteristics play important roles in rising air masses and in daily spatial distribution of the precipitations in complex region. As a result, its spatial discontinuity and behaviour in complex areas can affect the spatial distribution of precipitation. In this work, a two-fold concept was used to consider both spatial discontinuity and topographic and wind speed in average daily spatial precipitation estimation using Inverse Distance Weighting (IDW) and Multiple Linear Regression (MLR) in tropical climates. First, wet and dry days were identified by the two methods. Then the two models based on MLR (Model 1 and Model 2) were applied on wet days to estimate the precipitation using selected predictor variables. The models were applied for month wise, season wise and year wise daily averages separately during the study period. The study reveals that, Model 1 has been found to be the best in terms of categorical statistics, R2 values, bias and special distribution patterns. However, it was found that sets of different predictor variables dominates in different months, seasons and years. Furthermore, necessities of other data for further enhancement of the results were suggested.

中文翻译：

---

通过使用热带气候中的地形和风速变量，使用多重线性回归对平均日降水量的空间估计

复杂的地形和风特征在气团上升和复杂地区降水日空间分布中起着重要作用。因此，其在复杂地区的空间不连续性和行为会影响降水的空间分布。在这项工作中，在热带气候中使用反距离加权 (IDW) 和多元线性回归 (MLR) 估计平均每日空间降水时，使用了双重概念来考虑空间不连续性以及地形和风速。首先，通过两种方法确定干湿天数。然后，基于 MLR 的两个模型（模型 1 和模型 2）应用于雨天，以使用选定的预测变量来估计降水量。这些模型适用于每月明智的，在研究期间分别计算季节和年度的每日平均值。研究表明，模型 1 在分类统计、R2 值、偏差和特殊分布模式方面是最好的。然而，发现不同的预测变量集在不同的月份、季节和年份中占主导地位。此外，还提出了进一步增强结果的其他数据的必要性。