Prediction of Next Day Precipitation State by using Daily Temperature State over Iran

In this paper we attempt to investigate potentials of temperature estates for prediction of precipitation stance ‎based on conditional probability. For this purpose the daily precipitation and temperature of Iran data base are ‎used. This database has been established using 1436 stations for precipitation measurements as well as 664 ‎stations for temperature’s over Iran during1961/3/21 to 2004/12/31 and 15× 15 KM resolution. Accordingly it ‎has created 7187 pixels all over Iran. All calculations have been accomplished on these pixels. Probability ‎distributions of each pixel and their 25 and 75 percentiles of temperature have been estimated. Based on this ‎classification the cold, moderate and hot days have been defined as well as the no precipitation, low ‎precipitation and high precipitation days based on mean of probability distribution in each pixel. The join and ‎conditional probability have been calculated for every pixel to show every attitude of precipitation in compare to ‎one day previous. The results showed the precipitation – temperature relationship has tempo-spatial patterns ‎that cause very uncertainty in forecasting. However each day precipitation prediction based on previous day is ‎more probable in compare with later days. In some places such as southern coasts high precipitation is more ‎probable after a cold day. Meanwhile the dry days is more probable after a cold day in northwest of Iran as well ‎as in north of Hormoz strait, Khozestan coasts and Sistan region. No precipitation day after moderate day is ‎probable in 80% of times in south and north of Khozestan, from west of Hormoz strain to Baloochestan.‎