Evaluation of the benefits of using a backward chaining decision support expert system for local flood forecasting and warning. Issue 4 (26th January 2018)

• Record Type: Journal Article
• Authors: Zhang, Xiaoyin; Moynihan, Gary P.; Ernest, Andrew N. S.; Gutenson, Joseph L.
• Other Names: Cortez Paulo guestEditor.; Santos Manuel Filipe guestEditor.
• Is Part Of: Expert systems. Volume 35:Issue 4(2018)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2018-01-26
• Subjects: backward chaining -- expert system -- flood forecasting -- python; Expert systems (Computer science); 006.33
• DOI: 10.1111/exsy.12261
• Languages: English
• ISSNs: 0266-4720

Abstract

Abstract: Nationwide flood forecasting and warning are available through mass media. However, running the complex numerical models requires enormous computational resources. In addition, the comparatively low accuracy of prediction for a certain region such as a small town, a community, or a single house, causes false alarms and improper responses and thus the unnecessary loss of property and/or life. One potential solution to advance forecast accuracy without occupying substantial computational resources is to develop a stand‐alone local flood forecasting and warning expert system incorporating sufficient data on local hydraulic and hydrological factors and local historical experience. To date, there has been a limited amount of work developing expert systems in this area. In this paper, we discuss the development and implementation of an expert system for local flood forecasting and warning. With its extensible knowledge base combined with the information provided by users, this expert system provides reasoning routines and forecasting on the flood warning stages, possible consequences, and recommendations for community managers, landowners, or the public in general.