COMPARISON BETWEEN GPM AND GRIDDED PRECIPITATION PRODUCT FROM INTERPOLATED RAIN GAUGE MEASUREMENts

摘要

Due to sparsely distributed synoptic stations in the Philippines, representation of precipitation is inadequate. Satellite-derived products, such as the Tropical Rainfall Measuring Mission (TRMM) and Global Precipitation Measurement (GPM), are available to complement synoptic data for more precise and robust measurements. However, these products tend to over- or underestimate rainfall depending on the season. Precipitation measurements from a dense network of rain gauges in the Philippines are available courtesy of the Department of Science and Technology - Advance Science and Technology Institute (DOST-ASTI). This study aims to create a gridded precipitation data over the Philippines from interpolated rain gauge measurements and validate its performance through comparison with GPM. Comparison of rain gauge data with overlapping synoptic measurements in a 0.25° grid showed distance-dependent likelihood. Thus, inverse distance weighting (IDW) was perfonned to combine multiple rain gauge measurements representative of precipitation over 0.1° grid. Strong positive correlations (r = 0.58 to 0.97) were obtained from comparing IDW values and monthly synoptic measurements suggesting that rain gauge measurements may be used as complimentary ground data. IDW values were then subjected to kriging and iterative method to create a monthly gridded precipitation. Results showed that the interpolated precipitation follows the wet and dry season as well as monsoon seasonality in the Philippines. The calculated differences between the interpolated precipitation and GPM monthly product revealed that the interpolated precipitation have lower estimates with respect to GPM regardless of season. This is in contrast with the strong positive relationship between monthly synoptic measurements and GPM (r = 0.856). Low estimates from the interpolated precipitation is due to isolated zero measurements from various rain gauges. This may be attributed to maintenance issues of the sensors where the data do not report broken and erroneous instruments. Further filtering of the dataset and regular maintenance of the sensors must be done to better incorporate the use of dense rain gauge network for the creation of a Philippine gridded precipitation.