Skip to content. | Skip to navigation

 Back to Classic
Personal tools
You are here: GES DISC Home Precipitation Additional Features Science Focus TDST_SCI TRMM IMPROVES RAINFALL FORECASTING


Page Contents:


Tropical rainfall forecasting is a challenging task.It is common that rainfall forecasting for day one is reliable, but that forecasting for the second, third and more distant days in time, forecast reliability, which is a quantitative measurement of similarity between forecasts and observations, will degrade significantly. To improve tropical rainfall forecasting is important, particularly when it comes to hurricane tracks and rainfall accumulations. Moreover, the improvement will have an integrated effect on the model global scale behavior of rainfall, and help us understand the model rainfall patterns.

In order to improve fainfall forecasting, atmospheric scientists have concentrated on three main aspects: improving numerical weather forecast models; acquiring accurate observations for model comparison and validation; and developing analysis or initialization methods, which incorporate observational data into the model and give accurate initial conditions for the forecast.

Sample Application

One major accomplishment deriving from the use of TRMM satellite data is shown by theshort-term rainfall forecast study carried out at Florida State University.In this study, TRMM rainfall data and SSMI data are included in the multi-analysis process of the numerical weather forecast model.The results show that global, as well as regional forecasting reliability are enhanced by the addition of the TRMM data to the data pool.TRMM data are also used as the validation data for deriving the statistical parameters of the multi-model super ensemble system.The resulting forecast correctly predicted the tracks of major hurricanes in 1999 as well as dramatically increasing the accuracy of rainfall forecasting. Scientists attribute this success to a combination of improved analyses delivered by the super-ensemble approach and inclusion of accurate rainfall estimates over the tropics from the TRMM satellite.

The figures below exhibit a three-day rainfall forecast made by the multi-model super ensemble method.

3-day rainfall forecast of super
ensemble model

The top panel shows the initial precipitation field of the forecast model.The initialization process of the model makes use of the precipitation observed by TRMM TMI and SSMI (Special Sensor Microwave Imager) at 12Z Oct. 27, 1999.The heavy precipitation (red shading) along with supercyclone Orissa occurs over South Asia, over Bay of Bengal, and around Indonesia.The following three pairs of figures show the comparison between the forecast and the observed precipitation for the next three days.For day one, the forecast rainfall field seems identical to the observed in terms of rainfall regions and intensity.For the day two and day three, the predicted precipitation are still fairly close to the observation, but the discrepancies seem to develop around Orissa's center precipitation region.

The researchers at Florida State University currently extend their short-term rainfall forecast to thirty-day forecast experiment using TRMM and SSM/I data in near real time.For the detailed description about this study, please see TRMM site, then link to"Latest News", see released news on01-12-2000 . The detailed description of multi-model and multi-analysis forecast model are given by Simpson et al. and by Krishnamurti et al.


Simpson, J., Kummerow, C.D., Meneghini, R., Hou, A., Adler , R.F., Huffman, G., Barkstrom, B., Wielicki, B., Goodman, S.J., Christian, H., Kozu, T., Krishnamurti, T.N., Yang, S., Ferrier, B., 2000: The tropical rainfall measuring Mission (TRMM) progress Report, Earth Observation and Remote Sensing, Vol. 18, August Issue.

Krishnamurti, T.N., Kishtawal, C.M., Bachiochi, David, Zhang, Zhan, Larow, Timothy, Williford, Eric, Gadgil, Sulochana, and surendran, Sajani, 1999: Multi-Model Superensemble Forecasts for Weather and Seasonal Climate. Submitted to Meteorol. Atmos. Phys.


TRMM Data - 2A12

TRMM 2A12 are derived from TMI data.The data description and figures are provided on Cloud and Precipitation Formation topic of this page.

Relevant Links

Back to the Parent Page

Document Actions
NASA Logo -
NASA Privacy Policy and Important Notices
Last updated: Feb 10, 2016 05:14 PM ET