Utilization of Sentinel-2 imagery to identify a growth phase of rice plant in Cianjur Regency, West Java, Indonesia

As a tropical country, Indonesia has two seasons, where each season (rainy or dry seasons) lasts for 6 months. In Indonesia, the beginning of rice planting time occurs at the beginning of the rainy season, but a high diversity of rainfall causes the beginning of rice planting to become varies in some areas. In the area that have a same rainfall, however, the beginning of planting time can be also vary due to the rotation of irrigation water distribution. Seeing this phenomenon, the beginning of planting time is very important to be identified, since it can be used to estimate when the harvest time comes and also the production.The purpose of this study were to identify the growth phase of rice crops based on reflectance and vegetation index using Sentinel-2A imageries and distribution of growth phase of rice crops. The research method begins with a transformation of digital numbers into reflectance values in order to correct the effect of a light source, sensor, atmospheric. Analysis of reflectance (NIR) and vegetation index (NDVI, SAVI, EVI) values were carried out on 56 samples spread over five phases of rice growth, i.e. wet fallow, vegetative, generative, harvesting, dry fallow. The results showed that the NDVI can distinguish five phases of rice growth more clearly than EVI, SAVI, NIR. Pattern of reflectance value and vegetation index in the five phases of rice growth shows a quadratic model. Overall accuracy of the composite temporal of NDVI imageries is small (34,30%). The rice crops in harvesting phase is largest (42,04%)