|
|
|
Global Environment
|
Scaling of NDVI and VSW index between LANDSAT TM and NOAA AVHRR data
Mikio Sugita, Yoshifumi Yasuoka, Yoshiki Yamagata Yamagata and Masayuki Tamura
National Institute for Environemental Studies
Onogawa 16-2, Tsukuba, Ibaraki 305, Japan
Te; : +81-298-50-2589, Fax: +298-51-2572,
E-mail : sugita@nies.go.jp
Abstract
A scaling method was investigated to extrapolate the local information based on the fine spatial resolution data to more global scale by using coarse spatial resolution data covering wider area. In this case study, LANDSAT TM data was overlayed with NOAA AVHRR data, and the local landcover information derived from TM was extrapolated to continental scale with AVHRR data. As an efficient landcover property to correlate data from different spatial scale, the VSW index was applied which is a newly developed vegetation index reflecting the vegetation conditions as well as soil and water conditions. The VSW indices from both of TM and AVHRR data were compared and correlation between them was investigated. The result shows that the VSW indices from TM and AVHRR data correlated more than the NDVIs from the TM and AVHRR data.
1. Introduction
Deforestation, desertification and land degradation have been critical global environmental issues during the past decade. Monitoring and cover conditions and their changes is esential to the management of the environmental problems in both of the local and global scale. It is, however, no easy to monitor land conditions over large areas.
In the previous paper (Yasuila et al., 1995), we investigated a remote sensing method (scaling), which extrapolated the local information based on the fine spatial resolution data from LANDSAT TM to more global scale by using coarse spatial resolution data from NOAA AVHR covering wider area. The results indicated that pixel value of NOAA AVHRR data show linear relation with mixture condition among vegetation, soil and water from LANDSAT TM data. To find efficient properties useful for the scaling technique, it is necessary to examine which properties show evident relation when the properties are taken from sensors with different spatial resolution.
In this study, we applied scaling method to VSW index which is a newly developed index reflecting the vegetation conditions as well as soil and water conditions. The VSW indices from TM and AVHRR data were comapred and correlation between them was investigated taking area around Khon Kaen as a test site. In this paper, first, the VSW index is introduced and next, several property pairs from AVHRR and TM data are compared to find the efficient property for scaling.
2. VSW Index (Vegetation-Soil-Water Index)
Many vegetation indices, using red and infrared reflectance, have been devised to monitor land cover condition, especially vegetation. Ratio Index Normalized Difference Vegetation Index (NDVI) and Perendicular Vegetatin Index (PVI) are among the well-known indices of this type. The PVI (Richardson et al., 1977) was defined as the distance from the soil line on a scatter plot of near infrared (NRI) versus red reflectance (Fig. 1). The PVI was developed as a vegetation index to effectively (Fig. 1). The PVI was developed as a vegetation index to effectively monitor the vegetation biomass without affected by differences in soil index to effectively monitor the vegetation biomass without being affected by differences in soil background.
Fig. 1. Relationship between PV1 and the soil
line in a NIR-Red scatter plot
Yamagata et al. Devised a new index called a Vegetation-Soil -Water (VSW) index to monitor to monitor land cover conditions (Yamagata et al., in press). The VSW index is defined as a natural extension of PV1 for monitoring not only vegetation conditions but also soil and water conditions as well. The definition of VSW index is shown in Fig.2., which shows the relationship between VSW indices and the end member triangle on a NRI-Red Scatter Plot.
Fig. 2. Relationship between VSW indices and the end member traingle or NIR-Red Scatter plot.
The PVI measure only vegetation parameters, whereas the VSW index monitors vegetation, soil and water parameters simultaneously by measuring the distances PV, PS and PW vegetation water and soil, respectively.
|
|
|
|
|
|
|
