Abstract:

Aims There are problems with estimating vegetation cover using remotely sensed data. Many models have been developed by regression of field data and remotely sensed data, but this simple scale transformation often results in large errors. Our objective was to combine field data and multi-scale remotely sensed data to estimate vegetation cover for a typical temperate steppe of North China.

Methods Within our research area, we selected 49 sample fields from areas with high, medium and low vegetation cover and sampled each using 1 m plots nested within larger plots. We vertically photographed each 1 m sample plot with a digital camera positioned at 2 m height. We estimated vegetation cover in each image. Using these data and data obtained through ASTER and MODIS images, we developed a two-stage experiential model of vegetation cover based on the bottom-up method.

Important findings We accurately estimated vegetation cover of typical temperate steppe of North China at a regional scale based on our two-stage model using field data and ASTER and MODIS images. Using a series of MODIS images, it would be possible to estimate vegetation cover of typical steppe across China.

Key words: vegetation cover, grassland, field data, multi-scale remotely sensed data, two-stage experiential model