SPATIOTEMPORAL ANALYSIS AND MAPPING OF SULFATE DEPOSITION DATA OVER EASTERN USA

A spatiotemporal random field model is applied in the analysis and map ping of sulfate deposition over Eastern U.S.A. The integration of the spatial and temporal sulfate deposition variations introduced by the r andom field model has considerable advantages. It provides a rigorous representation of the spatially non-homogeneous/temporally non-station ary features of sulfate deposition. Deposition estimates and the assoc iated statistical errors are calculated at unmeasured locations/instan ts using a space/time mapping method. Composite spatiotemporal deposit ion maps are more accurate than purely spatial (kriging) maps. Moreove r, since a smaller number of monitoring stations are needed, the spati otemporal method may be cheaper to use than the spatial kriging analys is. Effects of the observation and mapping scales are discussed. The s patiotemporal model is general and can be applied to various data sets from atmospheric, environmental and meteorological sciences. (C) 1997 . Published by Elsevier Science Ltd.