B53C-0563
MINIMIZING GAPS OF DAILY NDVI MAP WITH GEOSTATIONARY SATELLITE REMOTE SENSING DATA
Friday, 18 December 2015
Poster Hall (Moscone South)
SeungJoon Lee1, Youngryel Ryu2,3 and Chongya Jiang2, (1)Seoul National University, Landscape Architecture and Rural Systems Engineering, Seoul, South Korea, (2)Research Institute of Agriculture and Life Sciences, Seoul National University, Seoul, South Korea, (3)Department of Landscape Architecture and Rural Systems Engineering, Seoul National University, Seoul, South Korea
Abstract:
Satellite based remote sensing has been used to monitor plant phenology. Numerous studies have generally utilized normalized difference vegetation index (NDVI) to quantify phenological patterns and changes in regional to the global scales. Obtaining the NDVI values during summer in East Asian Monsoon regions is important because most plants grow vigorously in this season. However, satellite derived NDVI data are error prone to clouds during most of the period. Various methods have attempted to reduce the effect of cloud in temporal and spatial NDVI monitoring; the fundamental solution is to have a large data pool that includes multiple images in short period and supplements NDVI values in same period. Multiple images of geostationary satellite in a day can be a method to expand the pool. In this study, we suggest an approach that minimizes data gaps in NDVI of the day through geostationary satellite derived NDVI composition. We acquired data from Geostationary Ocean Color Imager (GOCI) which is a satellite that was launched to monitor ocean around the Korean peninsula, China, Japan and Russia. The satellite observes eight times per day (09:00 - 16:00, every hour) at 500 x 500 m resolution from 2011 to 2015. GOCI red- and near infrared radiance was converted into surface reflectance by using 6S Radiative Transfer Model (6S). We calculated NDVI tiles for each of observed eight tiles per day and made one day NDVI through maximum-value composite method. We evaluated the composite GOCI derived NDVI by comparing with daily MODIS-derived NDVI (composited from MOD09GA and MYD09GA), 16-day Landsat 8-derived NDVI, and in-situ light emitting diode (LED) NDVI measurements at a homogeneous deciduous forest and rice paddy sites. We found that GOCI-derived NDVI maps revealed little data gaps compared to MODIS and Landsat, and GOCI derived NDVI time series were smoother than MODIS derived NDVI time series in summer. GOCI-derived NDVI agreed well with in-situ observations of NDVI using LED-sensors. We conclude that NDVI composition with geostationary satellite could effectively reduce data gaps in NDVI temporal series and spatial patterns.
