Analysis of rainfall data from the Tropical Rainfall Measuring Mission (TRMM).

摘要

The properties of tropical precipitation are investigated by estimating the probability density function (PDF) of rain rates using Tropical Rainfall Measuring Mission (TRMM) data. The gamma and lognormal distributions are utilized to fit the observed precipitation data. The minimum χ2 method is employed to a find a generalized PDF to represent each location and the resulting χ2 values are used to compare the performance. The two test functions both match the TRMM rain rates well, but the gamma outperforms the lognormal distribution in rainy regions, while the reverse is true in dry regions. Compared with the sample means, the parametric means from the fitted lognormal distributions overestimate and the gamma distributions underestimate. Constraining the shape of the distribution can decrease the difference between the sample means and the parametric mean from the gamma distribution significantly, although the χ2 values increase slightly.; Tropical wave characteristics are investigated using the TRMM rainfall and the outgoing longwave radiation (OLR) data. By using space-time cross-section analysis, lagged correlations, and space-time spectral analysis, wave modes are detected and their seasonal evolution and regional preferences are investigated. The Madden-Julian oscillation (MJO) and tropical depression (TD)-type disturbances are observed, as are waves that can be theoretically interpreted as linear equatorial waves. The cross-spectrum analysis indicates coherent wave variations in both OLR and rainfall fields. The asynoptic TRMM data may be contaminated by the diurnal cycle. Combining TRMM data with other satellites is desirable to avoid aliasing of the TRMM data.