This report describes the process of designing and building a prototype for visualizing the interpolated and height-adjusted heat sum on a web-based map in Agrometeorology Norway (LMT). Our system is designed for agricultural advisors in Norway, and the goal of the project is to enable the advisors to estimate the heat sum in places where there are no measurements of temperature. It is our hope, that the advisors will use the heat sum maps to convey climate statistics to their members.
We have used Brodersen’s design model to make the necessary decisions regarding the interface, content and functionality of the system. We have also constructed a flow chart to visualize the logical processes and dataflow. This chart was used as a basis for planning development activities and for internal discussions of the architecture and functionality of the system. Further, we have examined the data of LMT in order to find out if the density of the weather-stations is sufficient to carry out interpolation of the heat sum values between the stations. We have found that in most parts of Norway there is a sufficient weather-station density, but there are also places where only a height correction should be carried out. Through analysis of the existing station heights in LMT we have found a number of errors and missing data that need to be corrected before all weather-stations can be used in this system.
The goal of the project was to make the prototype based on LMT’s existing architecture and by using Open Source software. By searching different developer forums on the internet, we have identified technical components to carry out the individual parts of the process. In cooperation with NIBIO we have developed a webservice to carry out the extraction of data from LMT for use in the interpolation process. Interpolation and height correction will be performed using GRASS-algorithms, and we believe that the whole process can be automatized using PyWPS. The result of this will be a raster layer that must be coloured using a monochrome colour-scale, where colour density is used to differentiate between levels. The colourization will also make use of a normalized value range in order to compensate for a potentially large variation in the cell value ranges of the resulting rasters. The raster map will be presented as a webservice and will be shown in a map client based on Open Layers. In production we intend the map to be presented on the existing webpages in LMT.
The prototype we have made is not fully automated, but via the prototype we have been able to demonstrate that the combination of the chosen (or similar) components can produce the desired result. A test of the system showed that the users see a potential in the heat sum maps, and we have identified further development possibilities.
In the course of the project, we have adjusted goals and ambitions for the system and this report does not contain solutions for all the problems identified. Nevertheless, we have documented our work and our findings during the process, to pass on to NIBIO for use in further development.
We have learned from this process that this kind of development requires skilled developers. We have also learned that we can use our knowledge and experience in geoinformation together with acquired models and tools to participate in and contribute to the development of GIS components. We are able to analyze users’ needs, gain new knowledge and specify necessary development and adjustment, as well as to carry out tests to find out whether the developed functions have the desired effect for the users’ work process.