You should look at the changes in the frequency of extreme events as thunderstorms. If the frequency increased, this may explain the pattern. Thunderstorms can discharge larger amounts of rain in a short time period and in a small area (compared to rainstorms for example), thus, even if the amount of precipitation is decreasing in the whole area, it can increase at small scale because precipitation turn patchier. Increased frequency of extreme weather events is one of the main effects of global climate change. I think, nevertheless, that you can still explain the pattern if most precipitation comes from thunderstorms, even if the frequency is not increasing, just because statistics. At small spatial scale you will expect a strong variability in precipitation and, if your time window (the number of years that you are using to estimate ACTUAL precipitation) is small, then, given the strong variability, the estimation of local precipitation would be bad (I’m not saying that is wrong, just saying that, in statistical abstract terms, given the strong variability, the local rainfall estimation for a given year would be more variable than the historic rainfall estimation for the same region). Just imagine a matrix of 100*100 that explains precipitation in a region. Let’s say that the mean historic average precipitation of the area (the average of all cells in the matrix among all years) is 10 mm so the matrix that explains the historic (mean) rainfall is filled with 10’s. As the variability is high, the matrix that explain the rainfall in (lets say) 2012 is filled with 7000 cells with 0mm precipitation and 3000 cells of 20mm precipitation. The mean precipitation of the year would be 6 mm (the rainfall of the whole area decreased in 2012) but the rainfall in certain cells (certain areas) actually increased (1 mm versus 20 mm). Note that if every year the matrix behaves in similar way (most of the cells with low rainfall and a few with large rainfall) but the position of those cells just change, then you would expect a homogeneous mean historic precipitation (in the example, the historic mean matrix filled with 10 mm).

Dear Dr. Pedro Daleo,

Thank you very much for your clarification. Thank you for spending your valuable time in answering my query in detail.

Najim

This is the result of increased role of convective rainfall related to passing of ITCZ. It is evident that intermonsoon rainfall has been increased in both Dry and Wet Zones in Sri Lanka during the past 140 years (Jayasundara, 2004) You can check with the monthly contribution of rainfall in this patches. Pedro Daleo's conception conferms the evidance. Climate change has increased extreme events including thunder storms, which are charaterictics to the convective rainfall period when ITCZ passes through twice.

Ref. Sabaragamuwa University Journal, 2004

Well, Thunder storms are associated with convective process. In the tropical region, convection is one of the major atmospheric circulation processes which results the climate of the seasons. Therefore, there are two seasons in the tropics dominating convective process, when the region is under ITCZ. So, there are two seasons with domination conviction and there by thunder storms too. In contrary, during the other two seasons, Monsoon is dominating and no thunder storm activities possible as convection is suppressed.

Analysis of precipitation data related to Sri Lanka (separately for two regions – Dry Zone – Wet Zone) for the period from 1860 to 2000 reveals that precipitation in this two convection dominated seasons (1st and 2nd inter-monsoon) has been increased in both regions. Further, it is evident from many other published materials that due to increased energy in the atmosphere as a result of climate change, storm activities are increased, intensity of precipitation increased. Therefore, increased precipitation in the seasons is a result of strengthened convective process. Precipitation in the convective period receives as thunder storms. It has been increased.

Please see the publication - Method Title: Reason for changing rainfall patterns and increasing ...