Comparison of trmm3b42 v7 derived satellite rainfall with two rain gauge data for years 2000 - 2012, kilifi, Kenya

The study presents the results of rainfall data comparisons from three sources, namely the Kilifi Plantation Limited (KPL), Pwani University (PU) and Tropical Rainfall Measuring Mission’s (TRMM), TRMM3B42v7 satellite derived rainfall. The study aims at evaluating the ability of TRMM data to substitute local rain guage data as a sub-daily input parameter due to its finer temporal resolution. The study is motivated by the need to know the characteristics of sub-daily rainfall that would be useful in groundwater artificial recharge studies. The study’s methodology of comparing trend, seasonality and remainder of the rainfall data indicates that after removal of trend and seasonality in the rainfall data, the remainder signals’ cross correlation function between the datasets have a range of difference that may be accounted for by rainfall variability. The value for TRMM3B42v7/KPL is 0.43, for TRMM3B42v7/PU it is 0.48 and for PU/KPL is 0.49. This is further evidence of variability of rainfall in Kilifi as the cross correlation function of local ground gauges are not much different when compared with and to the satellite data-set. These Comparisons of similarities made on TRMM3b42v7 satellite derived rainfall against two local rain-gauges at a distance of 5 km apart from each other, was for data from 2000 – 2011. Out of the 4017 rainfall data events, 1801 are coincident (wet days) in at least one of the three rainfall data-sets. Kolmogorov-Smirnoff (KS) test that was used on the coincident rainfall events and had a p value of less than 0.05 for the two tailed test. Further cross correlation was done on the rainfall, which was decomposed for more analysis. The similarities obtained from the correlation of the decomposed trend, seasonality and remainder of the three datasets indicate that TRMM3B42v7 data can be used as a data input to model for the hydrological studies in the envisaged artificial aquifer recharge. This study also reveals the strengths and limitations of using the satellite derived rainfall product. The correlation of Kilifi rainfall data-sets with TRMM3B42v7 can be regarded as good when the cross correlation function between the ground gauges are noted to be in similar range to their correlation with TRMM3B42v7 data. This is despite close distance between the rain gauges.