AAIR Project No PL 94-2392

To progress Task 6, I propose to define a series of questions. I should like these questions considered by all members of the group, and the answers decided and agreed BEFORE the Umea meeting. From there I will propose a series of steps (to be broadcast well in advance of the meeting) for implementing these decisions, and I suggest that it is the details and timetabling of these steps which we discuss at Umea.

My strategy will be:
Pre-Umea:
To define the products that we will deliver at the end of the project.
In order to do this we must identify:

• For whom we are developing these products
  
• What the requirements of these "End Users" are.
  
• And, therefore, how we are going to satisfy these requirements
  

  At Umea:
  To agree a timetable for delivering these products (I will provide a suggested timetable and series of steps as a starter prior to the meeting).

  (n.b. I have just remembered something rather important that I didn't mention in the Brussels Meeting write-up, and that was that Dr Arabatzis was very keen to see growth models incorporated as a component, as a means of tying in the temporal aspects, and for future incorporation of economic estimates)

  Please bear in mind that this is entirely my interpretation of developments thus far. These suggestions are not necessarily the views of any other individual on the project. I am putting forward these ideas and suggestions merely as proposals and to intitiate discussion and debate. If you strongly disagree with any of these then I welcome any input, adjustments and additions (and deletions!)

  TASK 6 Questions

  The demonstrator has thus far provided a means of envisaging component and model integration, and given some insight into the apparent pitfalls. The approaches to addressing the issues of model and data applicability, validity and error have provided a framework for achieving integration. However, there are obvious limits, in terms of time, to the degree to which some of these issues can be pursued. It is also, perhaps, unrealistic to attempt to produce a complete single system which incorporates all the models and a single user- interface through which to directly access and run those models given the data. So the question has to be:

  EXACTLY what should the final product or products be at the end of the project?

  In order to answer this, it seems useful to try to define the "End Users" for whom the product(s) will be developed, to define their requirements, and therefore to develop an idea from there of exactly what the product(s) should be.

  In terms of defining "End Users" there appear to be two issues:

  a. What should the product looks like at the end of the project? In this case the "End Users" are considered as the EC, and Academia.

  b. What should the product aim to evolve towards in terms of Forest Management?

  Here the "End Users" being considered are individual forest managers, planners and perhaps government administrators.

  These groups have different requirements. Some aspects of these requirements can be satisfied by the project work, others cannot, and it is important to establish what these are.

  1. "End User" Requirements

  EU
  That we have fulfilled our promises:
  

  • I "to produce generic models which use factors common to wind, snow and fire damage.
    II "underpin those models with an understanding of the forces and site factors acting on single trees or the fuel hazard risks associated with site factors to produce risk assessments to forestry of minor or catastrophic damage.
    III "To test these models as a way to derive long term strategies for managing forests that optimise wood production.
    
  Academia
  
  • I Publications on findings
    II Workshops and presentations of findings (particularly with respect to the 1998 Conference)
    
  Forest Management
  A system which can be used to plan and manage future and existing forests, which will explain, for a given scenario, the risks and management options, within what limits and based on what assumptions. Consider some means of incorporating economics into the model (but not actually do this).
  (how do the requirements of different forest managers differ - for example operational foresters, to government administrators? Scale?)

  2. How can we address these needs?
  What should the "products" be?

  EU
  Addressing the needs:
  
  • 1. We need to define how generic our models are. Using the concepts of "data space" and "geographic space" we should map where models can be used across Europe
    2. We must define the limits of model applicability
    3. The scales and quality of data to which models can be applied must also be established, and the success of the models in terms of damage prediction must be defined through both error tracking facilities within the models and model validation.
    
  Presentation of the above information:
  We must produce some means of presenting the information from the above three points. This could be done via a series of maps and guidelines, but could also be incorporated in a package which is more demonstrative of model use with respect to questions that might be asked. This could be produced on the web using the framework suggested in Inveraray, whereby the "system" interprets a users query, and using the maps and guidelines of model applicability and data suitability, directs the user to the suite of models appropriate for the problem. However, this interface would not actually RUN any of these models.

  We should identify what questions this system would be answering and at what scale the problems will need to be addressed.

  I suggest there are three questions:

  • 1. "What parts of Europe would be suitable for planting a new forest in terms of damage risk?"
    2. "Given a particular unforested location, what silvicultural strategies should be adopted which would optimise wood production by best managing damage risk?"
    3. "Given a location with an existing forest, what silvicultural options exist which would minimise the risk of damage?"
    

  Products identified as "Academia" and "Forest Management" end user requirements will also address EU requirements.

  Academia
  Addressing the needs:
  

  • 1 Joensuu "STORMS" Conference
    2 Publications timetabled (?)
    3 Conferences and Journals to target
    4 Planning and running workshops for different aspects of the project
    
  Presentation of the above information:
  
  • I suggest these items fall within the remit of other discussions, and are less complicated for us to debate, so we should not spend time on them during the time allocated to task 6.
  Forest Management
  Addressing the needs:
  
  • A worked example of integrated tree and climate models which runs for a demonstration data set. This will be designed to demonstrate the input and output of the models and the meaning of the results in terms of forest management. A suite of worked examples should be put together so that every component developed in the project is incorporated in at least one of the worked examples. (For example some demonstration of Aberdeen's compression wood findings, or the implications of Galway's dynamic versus static tree loading results might, if demonstrated as a separate entity within a worked example, facilitate the understanding, by the forester, of the workings of the models, and instill an awareness of their usefulness and their limitations). This product is compatible with (and indeed the same as) the products which the Forestry Commission and (I believe) Erik are aiming for.

  Presentation of the above information:
  

  • PC or workstation-based developments (according to the models hardware/software requirements) with a basic user-interface which permits visualisation and some degree of interaction with the models and data. It should also incorporate a means of communicating errors and model limitations to the forester.

  Further suggestions?

  PRODUCT SUMMARY

  1. Web interface to explore component models in terms of user requirements (but not to provide data or actually run the models)
  2. Publications, workshops, and the STORMS Conference
  3. A package, or set of packages, which demonstrate the use of selected integrated models for the management of damage risk.
  4. Sets of maps and guidelines for the use/applicability/validity of models
  5. Information about the issues of scale and error in the models (quality of model output)
  6. Documentation of models and data used in the project
  

  3. What steps do we need to follow in order to complete these products by the end of the project?

  Immediate feedback required before 20th September 1996:

  1. Agreement that these are the products we must deliver

  2. Further suggestions/considerations/comments/refinements

  3. From every participant I would like:

  • a) A complete list of data and geographic factors which must be considered when looking at the applicability and limitations of all the models (for example, for the snow model:
    The model cannot be used outside the United Kingdom (because the relationships were developed using only UK data)
    The model was developed using data over the last 50 years, therefore considering return periods of more than 50 years may well produce unreliable results. The model is more reliable in areas which lie close to meteorological stations (list of stations))

    b) The scale/resolution for which the model works best (ie the scale/resolution of the different input parameters)

    This information is being compiled and made available through the metamodel forms. Therefore, I would ask for the completion of the forms by those who have not yet documented their models in this way, and for the checking of completed forms to ensure that they provide this information in full. I realise that many of the models are not yet fully developed, but some preliminary documentation should still be completed, and as the model is developed this can be updated. Participants should indicate on the form (in the "other information" box) whether the model is complete, and if not, what further developments it will undergo.

  4. Completion of "Results Issues" pages (which I will make available by the 30th August), for all the participants currently involved in the demonstrator. This means documentation of the results of the demonstrator. The results will be maps and tables which are the results of running the model on the data (I realise that in some instances this has not been possible), a detailed description of the problems which were encountered and some indication of whether the model-data interraction will be possible or not.

  5. Identification of the complete set of inter-participant interractions that are possible. (ie an extension of the demonstrator matrix to include all participants) I will make this large matrix available on the web (and in the post for non-webbers) within the first week of September using the information from the Inveraray meeting. This can then be changed and extended according to participants' ideas. I propose that further pursuit of the demonstrator would be a profitable means of continuing the process of identifying and demonstrating the model applicability, validity and data suitability issues.

  6. I should like all model developers to consider the possibilities for including error propagation and calculation facilities within their models (I am using the term "model" in the widest sense, to include the development of any method or relationships such as the fuel map calcualations by Portugal, and the wood properties and tree loading relationships of Aberdeen and Galway).

  7. I should also like participants to consider how the scale investigations might proceed for their models and data. (I will use the information from the metamodel forms to come up with some plans for this).

  8. Once the Product Definitions have been agreed (after 20th September) I will work out the steps we need to follow to develop and complete these products, and suggest a timetable for these developments which we will discuss in Umea.

  ---

  Private

  | Administration | Data | FTP site (MLURI) |

  ---

  Open

  | Description | Objective | Participants | Photo Gallery | WWW Resources | PROJECT home page |

  ---

  Alistair Law - a.law@macaulay.ac.uk
  
  Last modified: Tue Aug 13 11:54:08 BST 1996