Sentinel-1 Dual-Polarization Data Analysis to Identify Paddy Growth Stages in Indramayu District

Abstract

Abstract Sentinel satellite imagery using radar sensors for one particular area with the same orbit can be compared every 6 days and freely available to be downloaded. This advantage can be exploited to regularly identify land surface changes in a particular region. Paddy is a fast-growing crop with an approximately 120-day life cycle and specific growth stages related to phenological aspect, especially to changes in plant height. Hence, paddy growth stages can be monitored using this radar sentinel satellite with dual polarization data. Radar satellite is also cloud-free and therefore suitable for tropical regions such as Indonesia. This paper describes the results of a study on paddy field observation based on a correlation analysis between backscatter values of Sentinel-1A data with paddy growth stage observation. Sentinel-1A data from January 2018 to May 2018 were downloaded and processed using SNAP software. The study observed 5 stages: early vegetative (V1), late vegetative (V2), generative (G), Harvest (P), and Land Preparation (PL). An experiment was conducted by collocating the backscatter value of pixels at the location of the paddy growth stage observations, grouping and calculating the frequency for each class and then removing the class that had small frequency. These steps were carried out by way of iterations until the condition was above 5% of the threshold. Classification was performed based on range value obtained from the experiment to create paddy growth stage polygons. Correlation value between backscatter value σ0VH and paddy growth stage observation was 0.758. While correlation value between backscatter value σ0VV and paddy growth stage observation was 0.537.