No abstract

[1] In this study we have assessed the feasibility of a nonlinear technique based on genetic algorithm for the prediction of summer rainfall over India. The genetic algorithm finds the equations that best describe the temporal variations of the seasonal rainfall over India, and therefore enables the forecasting of the future rainfall. The forecast equation developed in the present study uses the monthly mean rainfall during June, July, and August for past years over five rainfall homogeneous zones of India to predict the seasonal rainfall (JJA combined) over the Indian landmass.

Wave prediction and hindcast studies are important in ocean engineering, coastal infrastructure development and management. In view of sparse and infrequent in-situ observations, model derived hindcast wave data can be used for the assessment of wave climate in offshore and coastal areas. In the present study, MIKE 21 SW Model has been used to carry out wave hindcast experiments in the Indian Ocean. Model runs have been made for the year 2005 using QuickSCAT scatterometer winds blended with ECMWF model winds. In order to study the impact of southern ocean swells, the model has been run in two different domains, with the southern boundary being shifted far south for the Domain 60S model. The model simulated wave parameters have been validated by comparing with buoy and altimeter data and various statistical yardsticks have been employed to quantify the validation. Possible reason for the poorer performance of the model in the Arabian Sea has also been pointed out.

A technique based on genetic algorithm (GA) is applied for predicting wind field in the Bay of Bengal (BOB) using satellite scatterometer observations. Empirical orthogonal function (EOF) analysis is used for compressing the spatial variability into a set of eigenmodes. The time series of each principal component (PC) is subjected to singular spectrum analysis (SSA) and GA is applied to the resulting filtered time series. The forecast PCs are weighted by the spatial eigenmodes for computing forecast wind fields. Predictions made up to 5 days in advance are found to be superior to forecast by persistence method.

The paper examines the variability of vertical humidity profiles over the Indian oceanic region using a set of 1200 radiosonde observations spanning 10 years (1982–1991). The examination is based upon the method of empirical orthogonal function (EOF) analysis. The first EOF explains 61% of the total variance and the first three EOFs together account for 85% of the total variability. The first principal component is almost perfectly correlated with the total precipitable water (TPW) and the second one is well correlated with the ratio of boundary layer moisture and TPW. This fact and an inequality derived from the analysis of the variances of individual terms of the EOF expansion of specific humidity are utilised to establish an algorithm for retrieving humidity profile from satellite microwave measurement of TPW over the region of study. Power of the retrieval technique is demonstrated using 127 independent radiosonde measurements and by plotting the profiles of rms error and bias. The method is found to be distinctly superior compared to a power‐law retrieval. A few examples of profile retrieval from satellite measurements of TPW have been checked against colocated radiosonde measurements. Some examples of retrieval show that the method is uniquely able to capture the humidity variability in the boundary layer, particularly the high moisture loading in the monsoon season.