The present study was undertaken with the main objectives of analysing, documenting the traditional means of weather forecasting followed by farmers; impact of weather forecasts in terms of decision making and extent of adoption by farmers; along with constraints and suggestions experienced by farmers in adopting the weather forecasts. A total of 180 farmers were selected randomly @60 each from Ananthapur, Rangareddy and Visakhapatnam districts of Andhra Pradesh state. Sixteen independent and two dependent variables (decision making behaviour and extent of adoption of weather forecasts) were selected for their relation analysis. Secondary data on rainfall, drought occurrence was considered for the study. Statistical procedures like frequency, mean, standard deviation, correlation, and regression were used for analysis.

About 66 bio and non-bio indicators for rain forecasting were documented from the farmers and were detailed with scientific explanation. The common bio indicators for rain prediction in the selected districts were movement of dragon flies, jumping cattle, poultry inserting feathers in the soil, flocking of sheep and goat and singing by black cuckoo bird. The common non-bio indicators were halo around the moon, formation of rainbow and red clouds at sunrise and sunset. Data on weather folklore and proverbs related to rain prediction were collected from farmers. Fellow farmers were the major source of information on weather forecasts available to the respondent farmers. Majority of them had medium decision making and extent of adoption of weather forecasts.

There was no significant difference among all the selected four almanacs in predicting rain for Ananthapur district. Rainfall prediction by almanac number four was found to be reliable for Rangareddy and Visakhapatnam districts. Drought prediction by almanac number one was accurate nine out of ten years for Ananthapur district. Almanac number four was accurate all ten years for Rangareddy and six years for Visakhapatnam respectively in predicting drought.

Correlation analysis revealed that knowledge on weather, innovativeness, scientific orientation, attitude towards weather forecasts, value and risk orientation had positive and significant relationship with both decision making behaviour and extent of adoption of the farmers.

Many farmers considered 'absence of location specific weather forecasts’ as major problem in adopting weather forecasts followed by 'poor reliability of weather forecasts’,’ failure of majority of weather forecasts’, and 'no proper extension service in weather forecasting.’ Farmer participation at community level should be integrated into information dissemination systems like media for realizing the maximum benefits of weather forecasts.

The study concludes that with ever changing weather patterns, proper recognition of bio and non-bio indicators of rain prediction is very essential. Adoption of weather forecasts shall increase if needs and preferences of farmers become an integral part of scientific weather forecasting. People associated with traditional and scientific weather forecasting should work in tandem rather than in isolation to solve the contemporary weather related problems confronted by the farmers.