ゼミ 2009/10/28

Cropland are discrimination using phenological change information

Background

◆Agricultural practices are continually changing at various spatial and temporal scales in response to local management decisions and environmental factors

◆Accurate cropland area extraction from satellite images poses problems due to wide difference of cropping culture

◆Detecting the situation of cropland is important for agricultural management, land degradation and ecosystem assessment

◆Monitoring of crop development and cropland area depends on several at cropping season

◆Time series NDVI, multi-temporal and seasonal images have been widely used because of its correlation with plant biomass and vegetation cover

◆Since most of the landmass covered by vegetation, the phenology based metrics will be more effective

Object

◆Discriminate cropland area from natural vegetated classes in arid and semi-arid region based on vegetation phenological change information

Study area

◆Geographic location:40N-60N,62d15E-141d35E

◆Countries:full of Mongolia, part of Russia, Kazakhstan, China, North Korea

◆Temperature:cold winter and hot summers have large daily and seasonal temperature ranges

◆Amount of rainfall is different, mostly have in summer

Land use characteristic

◆Major land cover types are forest, cropland, water, grassland, sparse vegetated and desert

◆Cropping system are mostly characterized as rain-fed cultivation and single crop per year

◆Growing season lasts from May until September

◆Main cultivated crops are wheat, barley, cotton, vegetables, fodder and rice

Used data

◆MODIS VI data from April-October, 2000

◆LANDSAT scenes from Maryland university

◆Global land use maps form IGBP-DISCover,GLC2000, Boston and Maryland university

◆Existing maps of cropland for each country

◆Google earth images

Why MOD13Q1 VI data?

・designed to provide consistent spatial and temporal comparisons of vegetation conditions

・

・

・

◆

◆

◆

◆

◆

◆

◆

◆

◆

◆

◆

◆

◆

◆

◆