本研究利用CSR提供的GRACE衛星觀測數據，其GSM版本的資料已扣除掉大氣和海洋模式訊號，只剩下陸地部份；資料所提供的球諧系數解最高至60階，相當於分辨率333公里左右。取最接近台灣中心的一點（E121°, N23°45'）為代表做解算，可將臺灣包覆在內。GRACE的重力資料變化主要與地表水循環有密切關係，透過地表水平衡方程式：ΔS/Δt=P-E-R，可以了解之間的關連性。ΔS為流域蓄水量變化，P為降雨量，使用中央氣象局的雨量資料透過權重解算得到；E為蒸發散量，使用Hamon（1979）公式透過溫度資料計算；R為逕流量，取自於水利署統計資料透過單位換算獲得。最後將 GRACE測量所得到的重力資料轉換成等效水深（ΔS），比較兩者之間的差異性，以了解GRACE衛星測量對於臺灣地區重力的精確度，進一步用來分析臺灣地表水循環現象。研究結果顯示GRACE衛星測量和P-E-R相關性不高，主因為臺灣地勢高聳、坡陡流急，致使雨水無法長時間停留在臺灣島內，GRACE衛星測量所得為月資料的變化，可能無法量測到短時間內水量改變產生的重力變化；以及P、E、R各自資料處理上所產生的誤差原因導致。In this study we use the GRACE satellite gravity data provided by CSR, of which the GSM version has already had the non-tidal atmosphere and oceans components removed. The highest spherical harmonic degree is 60, corresponding to around 333 kilometers of spatial resolution. We take the point (E121°, N23°45‘) near the central part of Taiwan to represent the whole Taiwan. The GRACE data of gravity changes have relationship with surface water cycle through the surface water balance equation：ΔS/Δt=P-E-R, where ΔS means change in storage converted from GRACE gravity data to the equivalent depth, P means precipitation, E means evapotranspiration, and R means runoff. For P we employ the data of precipitation from the Central Weather Bureau by means of Thiessen Polygons Method. E is modled via the Hamon (1979) algorithm using temperature data. R is obtained from the Water Resources Agency statistics. Comparing between ΔS and P-E-R can help understand the precision of the GRACE Satellite measurement for Taiwan, as well as analyze further the phenomena of surface water cycle in Taiwan. The results show that the correlation is low between the GRACE satellite measurements and the P-E-R. The main reason is Taiwan's high terrain, steep slope and rapid runoff, resulting rainfall can not stay long on land.
