Verification of Weather Predictions Using Voluntary Weather Observations Via WhatsApp and Google Forms During the Dry Season 2021

Abstract

The weather data that can be obtained through government institutions is very limited, whereas in order to increase the accuracy of weather predictions a homogeneous and dense distribution of data is needed. Therfore it is necessary to increase the data and the purpose of this research is to create a simple and effective way to encourage the number of weather observations in Indonesia through the STMKG Weather Care program. Forms that are made as easy as for respondents to understand, simple, and don't take the time. Developed using Google Form and distributed via the most popular social media today, namely WhatsApp. The test results showed that social media has the potential to be used to support voluntary weather data. The form made is sufficient so that the respondents make relatively few mistakes in terms of the main content of the form. Moreover, the mistakes that are often made by respondents include filling in ID, and typing sub-districts that require manual correction. Based on the results of voluntary observations spread in almost all provinces of Indonesia with the most incoming data coming from the provinces of Central Java and East Java. Based on the evaluation results of 4 months of testing, weather variations and their predictions can be identified with an accurate distribution, with an average accuracy of 0.79. Differences in methods used in verification may affect accuracy.