Relationship Between Time Series Models for Water Table Depth and Physical211 Information

摘要

The relationship between parameters of autoregressive moving average exogenous211u001evariable (ARMAX) models and physical information that can be derived from 211u001edatabases such as digital topographical maps, digital elevation maps and soil 211u001eprofile descriptions was investigated at 51 locations with open sandy soils in 211u001ethe Pleistocene part of the Netherlands. The ARMAX parameters appeared to be 211u001eweakly related with physical information. Water table depths predicted with ARMAX 211u001emodels which were guessed using physical information have large systematic errors 211u001ebut relatively small random errors. It is concluded that the relationships 211u001ebetween time series model parameters and physical information can be improved 211u001ewith respect to the modeling as well as the quality of the data used.