Untitled Document

Source of Error/diffculties in project:

The origin data was imported from ArcView shape files. In IDRISI, it does not inherent the spatial data from ArcView; therefore, IDRISI will assign its own ID to each polygon. It gives user a bit trouble to function the reclass because it is time consuming.

Error is inevitable along working on a model. Throughout the project, the major difficulty is the technical problem while i try to reclass polygons into unique id.

Because i cannot reclass the raster image, further research cannot be done.

I still do not understand the reason why it fails to reclass. I have the similar kind of project in my Geog 451 course, and it works fine. However, we may still find some error in IDRISI even though we reclass and import correctly. According to our Geog 451 project, which we were working as a team, Tim Koh, Kim, Sachiko, Uncertainty and error are also found. I am not allowed to post any images in here because i have to respect the copyright issue. It is a group project; therefore, copyright is belong to the whole group.

Below section are showing error/difficulties that we may find if we keep on building the model, all error are based on assumption by my knowledge and experience in my Geog 451 project.

Error 1: Factors, fuzzy setting, and weight are subjective due to creater own interest. It will inherent error in the result because producer may have bias on several condition.

Error 2: Generalization on category. For example, Low density residential includes Mid density residential, institutional included, school, municipal recreation facilities. If it has a proper way to define the category, more precise and accurate result can be given.

Error 3: Edge Effect. Along the boundary of Coquitlam, high residential or low residential area take place. However it may not result as a growth in the future because it does not have open area or not meet the factors. It is not the case because i did not included Maple Ridge, Port Moody, and Port Coquitlam in the map. If adjancent cities are included in image, values along the edge will change.

Because City of Coquitlam does not have the latest digital data of land use, i cannot perform the CA_MARKOV module. CA_MARKOV needs two different years layer as input to calculate the future growth. I only have year 2001 data; therefore, cannot run the Cellular Automata Model.

Difficulties:

I. Cannot perform reclass in IDRISI.

II. IDRISI require more step to obtain the correct format of data, and time consuming.

III. Only 2001 data was used, not enough to build Cellular Model. Can only perform the MCE.

At first, i was choosing Hong Kong SARS as my project topic; however, i am not able to find much data in the topic. I got basic map, which has utilites, roads, boundary only. SARS is a very interesting topic because it may help to figure out the disease pattern, and useful in future studies. Hope I can have chance to do a research on it in future!

By Tim, 200021477

Nov 24, 2003