Perbaikan Distribusi Anomali Free-air di Pulau Jawa dengan Metode Kolokasi Kuadrat Terkecil.

Studies in geoid-gravimetric modeling conclude that one of the problems in precise geoid modeling in Indonesia is the condition of the gravity data. The main problem within the data is its distribution and the data quality. Alternatively, the data improvement can be done by prediction at location that have not been surveyed. Prediction makes possibility to rearrange a better data distribution and quality. Prediction on the gravity anomaly data was done by least squares collocation. The earth geopotential model EGM96 was used as a systematic part in the collocation. The covariance function of the gravity anomaly was derived from the Hirvonen function. The result from the prediction was evaluated by empirical testing with the calculation of difference values between the data observation and the prediction. The prediction result is free-air anomaly grided in 2'x2' at the Java island. The empirical testing on the free-air anomaly gives the average differences -0.007mGal with standard deviation 5.948mGal. This result is better than the existing data.Kata Kunci: Anomali gayaberat, kolokasi kuadrat terkecil, geoid-gravimetri