Premium
Inverse modelling of environmental pollution: The role of statistics
Author(s) -
Dovì V. G.
Publication year - 1991
Publication title -
environmetrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.68
H-Index - 58
eISSN - 1099-095X
pISSN - 1180-4009
DOI - 10.1002/env.3770020307
Subject(s) - inverse problem , computer science , regularization (linguistics) , inverse , operator (biology) , econometrics , pollution , mathematical optimization , mathematics , artificial intelligence , ecology , mathematical analysis , biochemistry , chemistry , geometry , repressor , biology , transcription factor , gene
Modelling is currently used in problems connected with the assessment of environmental impact. This class of problems is generally well defined, in that all the characteristics (such as location, strength, etc.) of potential sources of pollution are known. On the other hand, receptor analysis uses the experimental information available in the environment (such as air or water quality) to identify polluting sources. This is why they are generally called inverse problems. The reasons for the negligible amount of research effort in the application of inverse modelling is due to the ill‐posed character (i.e. lack of continuity) of the operator that describes the problem. Nevertheless, large strides have been made in the so‐called regularization of inverse operators, which makes use of additional available information. The aim of this paper is to explore how the statistical information employed in receptor analysis can be used for regularization purposes in the general inverse modelling of environmental problems.