Indigenous Knowledge Mobile Based Application that Quantifies Farmers’ Season Predictions with the Help of Scientific Knowledge

Abstract

This paper presents the development of the indigenous knowledge (IK) mobile based application that quantifies farmers’ season predictions with the help of weather data, satellite imagery data and the Single Exponential Smoothing (SES) model. The system facilitates the indigenous knowledge indicators’ collection and processing to compute farmers’ certainty level of an oncoming rainy season behaviour which is mainly categorized into abundant, droughts and floods. Despite the value of IK indicators, they can only tell the behaviour of the season without stressing on valuable scientific information such as rains onset, distribution, magnitude and cessation. To solve this problem, the researcher integrates the use of IK indicators with farmers’ historic data of periods when they have experienced abundant(normal) rains, less rains (below normal) and excessive rains (above normal). For each period, weather data (rains and average temperature) and satellite imagery data were collected, processed and stored in the database for use by the application. For each satellite image, the following land cover features: vegetation, soil moisture and waterbodies area cover were computed. The indigenous knowledge indicators were also systematically structured to enable certainty level computation of an oncoming season. Based on farmers’ predictions of the oncoming season behaviour, the system extracts historic data with respect to the predicted season and send the data to the SES model. The SES model will predict the next season’s weather data, vegetation, soil moisture and waterbodies cover data to help the farmer to have robust knowledge on the possible season outcome. The data is also downscaled to provide meaning to the farmers.