Methods for Comparing Global Vegetation Maps

Monserud, R.A. (1990). Methods for Comparing Global Vegetation Maps. IIASA Working Paper. IIASA, Laxenburg, Austria: WP-90-040

Preview

Text WP-90-040.pdf Download (1MB) | Preview

Abstract

Objective statistical methods are presented for comparing global vegetation maps. The methods are illustrated by comparing maps resulting from applying a modified Holdridge plant/climate hypothesis to various global climate projections and to current vegetation (the baseline). Five general circulation model projections (GFDL, GFDL-Qflux, GISS, OSU, UKMO) of expected climate resulting from doubling current CO2 levels were used as input to the modified Holdridge model. The Kappa statistic proved to be a useful and straightforward measure of agreement between maps. Furthermore, individual kappa statistics for comparing a given vegetation zone between two maps clearly indicated differences and similarities between maps. Additional summary statistics compare the change in area, latitude, and longitude between maps for each vegetation zone, as well as the distance and direction that each vegetation zone has shifted.

Actions (login required)

View Item