Abstract
An approach to use satellite data for quantitative estimation of the fixation of nitrogen by different vegetation types over river basin ecosystems was proposed and applied to a mountainous area located in the upper reach of Nagara river basin (Gifu Prefecture, Japan) covered predominantly with multiple vegetation types. In addition to satellite data, this approach also involved the use of census data, statistics data and previous results available in literature as supporting materials. With this approach, the annual amounts of nitrogen absorbed by different categories of forest vegetation classified in terms of species and ages using the satellite data were estimated. Validation tests were performed by comparing the satellite data-based estimations (Ndsatellite) with estimations from a budget model (Ndmodel) and statistics data (Ndrecord), respectively. It was found that the estimated NdsatelliteN (100.8-111.7kg ha-1y-1 for broadleaved forest, 35.4-71.4kg ha-1y-1 for evergreen coniferous forest, 91.4kg ha-1y-1 for mixed forest) coincided well with the values of Ndmodel (96.0-100.4kg ha-1y-1 for broadleaved forest, 48.0-74.0kg ha-1y-1 for evergreen coniferous forest, 82.3kg ha-1y-1 for mixed forest) and were somewhat larger than those of Ndrecord although the values of Ndsatellite and Ndrecord were all in the same order of magnitude for respective vegetation types classified.
