Shikher-Today

I began my day by resuming the realignment of population-density ASCII file which was 20% done on Monday. It took me approx. 3-3.5 hours to realign approx 80% of the file. The some error occurred, application closed suddenly and I lost all that data.

So, moral of the story that I spent approx 5 hours in that file and finally nothing in hand.

Then instead of restarting the process, I tried to find the solution through shell scripting to convert comma delimited file to space delimited. Fortunately I found one and then I converted that file into realigned ASCII file in 5 minutes (:D) through Unix Command.

Now, we have all the aligned data-set for approx 230 m cell size. On one computer I am creating the final dataset for GARP and its still under process. It may take more 30-40 minutes but I wont be here by then so I have left instruction to save that data, once processed by Desktop Garp Manager, on Sid's computer (as I won't be here tomorrow).

I also started the GARP model with 6 files (of 230 m cell size) approx. 40 minutes ago and its still running. It has not given even a single result out of 5 results. It creates a doubt. I am not sure what will happen when I will run the model with 23 files and how much time its going to take.

Shikher-Today

I began my day with data processing and then discussed about video with the team. After that I wasted 1-2 hours because of the abnormal behavior of ArcGis. Finally, I started using James' computer and it was working fine there. I think I shall have to use some other computer from now onwards, because we were facing the same problem in the last term. I don't know why it happens sometimes with Arc on my computer.

I have started re-sampling of data to approx 230 m resolution. Desktop GARP takes approx 3 minutes per re-sampled ASCII file to convert into RAW file, so its not that bad. Arc takes 15-20 minutes to give resampled ASCII file.

If would be good if we can clip some more area. I will talk about this on Friday. I have created tool in ArcGis to perform re-sampling and conversion to ASCII in single step.

I am running test-case with two re-sampled layers in GARP. It took approx. 10 minutes to produce one probability map with 2 input layers, still 4 more to go. So I think it won't be bad with 10-20 probability maps and 23 input layers.

Shikher-Today

I did following tasks today:

• Rectified the misalignment problem with the population desity ASCII file.
• Resampled the land cover data according to other parameters.
• Ran the GARP model with minimum option to test. It was successful.
• Now, I am running the GARP model with standard options for one species. Its still running. It took 3 hours to finish 33% of simulation.
• Worked on Tech Paper.

We are currently runnig a model for approximately 1 Km resolution data. I resampled land cover data to 1 Km resolution to meet GARP input requirements. Now I am thinking of making inputs with finer resolution like 200 m, 300 m, etc to get more accurate result. I will discuss with other team members about this on Friday and then I will start preparing new dataset with finer resolution.

Good news is that now everything looks fine and I don't think there is any problem with the functioning of Desktop GARP. Now, we can experiment as much as we want with the dataset like:

• Trying different resolution dataset.
• Trying different variables.
• Trying different study area.
• Trying different species.
• Trying various option given in Desktop GRAP.