IEEE Computational Intelligence Society
Intelligent Systems Applications Technical Committee (ISATC)

Taskforce "Earth and Environmental Sciences" (TF EES)

This Taskforce was established in August 2006 as a result of the decisions made by the governing bodies of the IEEE Computational Intelligence Society (IEEE-CIS). It works in close cooperation with the INNS Special Interest Group on Computational Intelligence in Earth and Environmental Sciences.

Background

Apart from the so-called process models (based on the detailed descriptions of the physics of the modeled process), the so-called CI methodologies become more and more popular among practitioners. CI methodologies include data-driven models that use the methods of machine learning (and, widely, computational intelligence) and hybrid models that combine components based on process models with data-driven components. In earth and environmental sciences hundreds of successful applications of CI approaches are known. The predictive methods used for forecasting different environmental variables (flows, water levels, currents, algae growth, etc.) employ connectionists methods, like neural networks, but also other methods, like SVMs, fuzzy rule based systems, model trees, etc., being often combined in committee machines and mixture models, or hybridized. Methods used for analyzing very large data sets and time series use various versions of PCA, ICA, wavelets, etc. Evolutionary and other random search techniques can be used for models optimization.

Application areas of CI methods include, but are not limited to:

• climate modeling

• weather predictions

• other meteorological and oceanographic applications

• geophysical data processing

• water resources and hydrology.

These CI applications are quite different, but there is a number of common elements related to the modeling methodology that are present in many applications.

Objectives

The objectives if this Taskforce are:

· to encourage and promote the responsible use of CI methods and methodologies developed in the research areas covered by IEEE-CIS in earth and environmental sciences, in academia, governmental agencies and consulting companies;

· to encourage members of IEEE-CIS to develop and apply methods that match the complexity and specific characteristics of problems encountered within earth and environmental sciences;

· to interact with other IEEE-CIS committees and groups in the matters related to the earth and environmental applications;

· to organize regular special session at the IJCNN, WCCI and other conferences and forums;

· to maintain links with the Technical Committees existing in other scientific communities and associations. In the first place, the links have or will be established with:

· Committee on Artificial Intelligence Applications to Environmental Science (American Meteorological Society, AMS)

· Sub-division on Hydroinformatics (European Geosciences Union, EGU)

· Joint committee on Hydroinformatics (IAHR/IWA/IAHS)

Members

Name	Affiliation	Email
Vladimir Cherkassky (chairman)	University of Minnesota, USA	cherkass [at] ece.umn.edu
William Hsieh	University of British Columbia, Canada	whsieh [at] eos.ubc.ca
Vladimir Krasnopolsky	Environmental Modeling Center, USA	vladimir.krasnopolsky [at] noaa.gov
Dimitri Solomatine	UNESCO-IHE Institute for Water Education, The Netherlands	d.solomatine [at] unesco-ihe.org
Julio Valdes	National Research Council, Canada	julio.valdes [at] nrc-cnrc.gc.ca

Recent Activities

World Congress on Computational Intelligence (International Joint Conference on Neural Networks, IJCNN), July 16- July 21, 2006, Vancouver, Canada: Special Session “Computational intelligence in Earth and environmental sciences ” (24 papers)

Mini-Symposium “Computational intelligence in water and environment”, together with the IEEE Computational Intelligence Society (Benelux Chapter), INNS SIG "CI in Earth and Environmental Sciences, and the Belgian-Dutch Society of Ecological Modelling, 15 December 2006, Delft, The Netherlands.

International Joint Conference on Neural Networks (IJCNN), July 31- August 4, 2005, Montreal, Canada: Special Session “Applications of Learning and Data-Driven Methods to Earth Sciences and Climate Modeling” (26 papers)