Characteristics

The proposed EMS consists of a web-based system that is accessible by anyone and is permanently connected to all relevant data streams. In addition to data, the web site contains analysis software, documentation for this software, analysis output, and references to literature pertaining to the ecosystem being managed. All of this content can be freely downloaded without restriction.

Essential abilities and characteristics of this EMS are as follows.

(a)

An intergovernmental agency is created and funded by the ecosystem-affecting countries (hereafter regulatory agency). This regulatory agency is responsible for maintaining at the web site, all source code, a user's manual, and data base documentation for the EMS. To ensure unlimited access, none the EMS source code is written in a proprietary computer language.

(b)

As will be explained in Section 2.2, the probabilistic ecosystem model is organized as an influence diagram. This model is multivariate, spatio-temporal, and is easily modified to include a wide selection of cause and effect relationships between independent and dependent ecosystem variables. Impacts of different management options on each country's economy are also included. A panel of scientists and business leaders build the initial influence diagram. Parameters are periodically updated via a statistical method that utilizes both monitoring network data and expert opinion.

(c)

The EMS can compute estimates of ecosystem and economy state variables at the current time point or future time points. Estimates of the state of both the ecosystem and the economy are delivered in two forms. First, a standard set of current and future state variables are automatically re-computed daily and are available at the EMS web site. Second, custom estimates and forecasts can be easily requested at the web site and a report of the results is automatically written for the requester.

(d)

By performing monitoring network redesign computations (e.g. Haas (1992)), the system automatically and periodically recommends modifications to monitoring networks to which it is connected. Because the system is gathering data from multiple countries, redesign recommendations are not biased by examination of only one country's monitoring data.

(e)

Advances in ecosystem models, econometrics, and statistical methodology are continually being made. To allow the EMS to be modified as such improvements are discovered and to have an organized way to select among competing models and/or statistical methods, the following protocol is followed. First, anyone with access to the web site can propose an alternate model and/or statistical method. Then, a competition between this proposed model and/or method against the official model and/or method is held by the regulatory agency. This competition entails first entering the competing model and/or method into the EMS and then finding which of these two models and/or methods gives the lowest out-of-sample prediction error rate on state variables as determined by a statistical hypothesis test. The winning model and/or method becomes the official EMS model and/or method.

The hope is that the use of such an EMS would raise the quality of policy creation discourse between public and private sectors because (a) all groups would use the same data and ecosystem-economy model in debates over which ecosystem management policy should be implemented, and (b) the debate would center around common, shared EMS-generated predictions of a proposed policy's impact on both the ecosystem and each country's economy.