Abstract:

Records of extreme precipitation were investigated using the Discrete Autoregressive Moving Average (DARMA) process, which can explain long persistences of wet and dry spells that exist in daily precipitation data. The results show that the daily precipitation with strong autocorrelation is inclined to be better fit by a Discrete Autoregressive (DAR) model. On the other hand, those data with weak autocorrelations tend to be best fit by a Discrete Moving Average (DMA) model. It can also be concluded that based on the records from extremely wet and dry regions there is no geographic preference regarding the selection of the best model. (KEY TERMS: extreme precipitation; flood; drought; autocorrelation; autoregressive; moving average.)

1. Paper No. 88106 of the Water Resources Bulletin.

2. Professor, Department of Civil Engineering, Ohio University, Athens, OH 45701.




For more information on this or any other paper contact Terry Chang by email at tjchang@bobcat.ent.ohiou.edu

Copyright © 1997