ENVS10006 Mapping Environments

Assignment 3: Digitising Change

He Wang (Jack) 1081155

Lyn Hirai Gray 1171207

Nina Bron 1277409

Assignment 3: Digitising Change

Aim:

To learn how to create spatial data from scratch using digitising techniques. And work with
aerial images from Web Map Services (WMS) to analyse any changes over time.

Task:

To analyse the area damaged by bush fires in Harrington, NSW.

1. Find the difference between the forest cover before and after the bushfires
2. Find the percentage of the municipality land damaged by fires
3. Find how many trees were lost
4. Find how many houses were affected
5. Find how close/far from settlements were the bushfires
Procedures:

Connect WMS service with QGIS:

1. Go ahead to https://www.nearmap.com/, login and click on ‘My API Keys’ on the

welcome page.
2. Open the service section, click on ‘Add Custom Service’ to add a service to explore
Harrington town in NSW.
3. Enter a recognisable name for the service.
4. Click ‘ADD AREA’ and look for Harrington town in the search box, then click
‘Harrington, New South Wales, Australia’ to navigate to it.
5. Include all necessary areas by drawing a polygon on the map, it can be helpful to
check its boundaries on a mapping service.
6. Type ‘Harrington, NSW’ in ‘Name’. Leave the dates blank. Click Continue.
7. If ‘Harrington, NSW’ appears, close the window.
8. Change ‘Choose application’ to ‘UniMelb API (Student)’ (or maybe ‘UniMelb API
(Stuff)’). Copy URL.
9. Open QGIS, connect the WMS service by right clicking ‘WMS/WMTS’ in the
Browser panel, select ‘New Connection’.
10. Name the WMS service and paste below the URL. Click OK.

Locate ‘Harrington 2427, NSW’ in QGIS:

11. Locate Harrington town in QGIS via WMS service. Choose ‘WMS/WMTS’ >
‘Harrington NSW’ > ‘Harrington NSW’ > ‘Harrington NSW’ > 2019-07-xx Taree.
12. Download Zip file ‘Postal Areas ASGS Ed 2016 Digital Boundaries in ESRI shapefile
Format’. Make sure the correct ‘shapefile format’ is downloaded, or it will not work.
13. Unzip the file into a created project folder.
14. Locate the project folder in the ‘Browser’ panel, open the shapefile (in .shp format). It
will create a new layer in the ‘Layer’ panel.
15. Right click on the new panel, choose symbology and unfill it, which allows to see
only boundaries.
16. It will be great to save the work now.

Digitising:

17. Digitising in both 1:2500, 1:10000 and 1:25000 scales. Find scale in the bottom panel,
enter the scale wanted and press the ‘lock’ icon to lock the scale.
18. Create a new shapefile layer by clicking the ‘shapefile’ icon on the left side bar. Give
the layer a recognisable name such as ‘25000before’. Choose the format as ‘.shp’.
Click the three dot icon next to the file name to choose a place to store the shapefile.
Change ‘File encoding’ to ‘UTF-8’. Change Geometry type to ‘Polygon’.
19. Click on the new layer, then click on the ‘Toggle edit’ button to edit the layer.
 20. Click on the ‘Add Polygon Feature’ icon on the top panel. Then begin to digitise by
left clicking and finish the shape by right clicking. Make sure all areas wanted have
been covered with polygon. Give the shape an ID when finishing the shape.

Calculating areas before, after and their difference:

21. Check the map measurement units. (Project > Properties > General > Measurements >
Set the units for distance measurement to ‘Kilometers’ and units for area
measurement to ‘Square kilometers’ and press OK.
22. Calculate the difference between before and after areas mathematically. Find the
attribute table of layers. (Right click on the digitized layer > Open Attribute Table)
23. Open ‘Field Calculator’.
24. Read the dialogue and check ‘Create New Field’, Name: Area, Type: Decimal number
(real), field length: 10, Precision: 5. And build an expression using the tools in the
middle column: Select Geometry and $area. Click Ok and the result will appear in a
table.

Geoprocessing approach:

25. Find a ‘Intersect’ tool. (Vector > Geoprocessing Tools > Intersection.
26. Select the Before Layer as an Input layer, and After layers as overlay layer. Make sure
the output is stored in the project folder. (Use the three dots button)
27. Calculate the area of the intersection layer, same way as Step 24.
28. Compare the results.
Results:

Scale Before After

1:25000 (Nina)

Total Area: 28.28 km 2 Total Area: 5.38 km 2

1:10000 (Jack)

Total area: 5.71 km2

Ps. the red area represent the
Total area: 29.32 km2 bushfire area

1:2500 (Lyn)

Total area: 29.02 km2 Total area: 5.67 km2

Analysis
The main difference between the forested area before and after as seen in the above table and
the map below:

The main difference is that there is a significant loss of trees, with a loss of 80.46% of trees,
or approximately 669,864 trees lost.

There were not any houses lost or damaged in Harrington. The closest what appears to be a
house was 1.5 m away, in the below picture:

Discussions:

Discussion 1:
Without outlining the methods used in this project, how else can the above issues be solved?

A possible alternate method is to compare photos on ground of before and after the fire to
estimate the affected area, however the coordinates of these photos would be needed. Another
way would be to highlight the infrared reflected by the trees for the before and after of the
bushfire. This is assuming the camera kept the infrared spectrum.

A method to estimate the number of trees burnt would be to count manually or get the
average amount of trees per km and multiply it to the closest km square of the affected area.
However, counting manually will take too long so estimating through an average would be
more practical. Another method would be to check on the Bureau of Statistics and apply it to
Harrington, which can also work for houses lost in the area as well.

By using photos taken around the edge of the burnt area, it can be used to determine which
settlement(s) are closest using Nearmaps.

Discussion 2:

- Can you describe the effect of different scale resolutions on the quality of your
digitised polygons?
● What are the limitations and benefits of being too zoomed out (small scale)?
● What are the limitations and benefits of being too zoomed in (large scale)
- What scale would you suggest is the best in this case?
- How is this process of abstraction with less detail called?

It is a lot easier to get the details right when zoomed in so a more accurate polygon to cover
the areas can be made. A disadvantage of being too zoomed in is that the polygon will need
to be made in small parts, as there did not seem to be a way to undo the creation of a point for
a polygon. Another disadvantage I noticed when making the polygons was that because many
details can be seen, it made it difficult to tell if something was not a tree or if it just had a
strong shadow over it.

Discussion 3:

- What are the benefits of calculating the difference without the use of GIS tools?
- What are the benefits of using the Intersection function?
- Which method do you think is more suitable for this task and why?
- One of your questions in Discussion 1 asked about percentages of the land being
affected. Describe how and perform this task.

Benefits of calculating without GIS tools is that if there is an error with how the location is
bound, it could make the area extremely large or small. It could also be advantageous to use a
different program if you are not used to using the GIS software. Such as just using Excel to
calculate the total area instead of using QGIS to do it for you.

The benefit to finding the intersection is to find any overlap between the areas in question.
I think the first method was the best, as it was straight forward to calculate the total area of
before and after, then finding the difference. The intersection method however, I did not think
it was very useful for this task as the intersection was usually just the area after the bushfire.
So it did not give any new information.

The percentage of forest land affected was calculated by first taking the difference between
the area of forest land before and after the bushfire, divide it by the area before the bushfire
and multiply it by 100 for the percentage:

(29.02-5.67)/29.02 *100=80.46%

Discussion 4:

- What issues did your group experience with the following:

● Nearmap and NSW Spatial Services WMS
● Digitisation at different scale levels
● Analysing before and after shapefiles of the forest

There was some difficulty with linking Nearmaps with WMS as the link address would have
the code in it so it was sometimes difficult to locate it or the textbox to type the code would
sometimes be greyed out.

Digitising at different scales created mostly time and consistency issues, the more “zoomed
in” your scale was, the more accurate your results, however this was time consuming. These
created inconsistencies between individual results as one person included a section they
thought was forest whilst another did not. However, this issue was not very problematic as
overall, results were not drastically different to one another.

Analysing the shapefiles of the forest before and after the fires did not prove to be a challenge
for our group.

Discussion 5:

- How did you find working with WMS and digitising? Was it intuitive? How can it be
improved?

Lyn: Using WMS and linking was not very intuitive, as a third party link was required and
would need research or prior knowledge to do so. Even then, a couple members of our group
had trouble with it. A way to improve this might be to be able to select a location from QGIS
rather than needing a third party source, create a link and link it to QGIS.

The digitisation process was more intuitive and simpler as the tools necessary were all in
QGIS and each were labeled what they did. Some prior knowledge was very helpful, but
since the polygon tool was similar to making shapes on Microsoft Word and other programs
it would not have taken many experiments to understand it.
Jack: In my opinion, there are both good and bad aspects of using WMS. Firstly, I am amazed
when I firstly operate on the linking, because theoretically a higher quality can be achieved.
With WMS, we can relatively easily build a project to compare a specific location two times.
At the same time, the final result will appear in higher image quality. I would not spare my
praise for those benefits.

But, it is quite not intuitive and easy to operate. In my experience, I have followed the
instructions provided, but when I connected my QGIS with WMS, there was nothing
appearing. To make sure if I did anything wrong, I have tried the exact same operation on a
university PC. I got what I want on my PC. So basically, there is a limitation of the operating
device for WMS. I would also say, photoshop would be very helpful when I get the result.

Nina: Personally, I enjoyed using WMS, I found the maps of a higher quality and it was
relatively intuitive. Because one could simply enter in a suburb or a postcode and it could be
located, this was invaluable for the kind of project we were doing. WMS is very powerful and
I liked being able to imbed the WMS features into the QGIS software.

The digitising process was even easier than I anticipated, having already had previous
experience making shapefiles on QGIS (thanks to assignment #2), I was able to digitise very
quickly. I found the tools were well labeled and with some experimentation I could digitise
the forested area accurately and quickly. However, it was a hassle to edit polygons if you had
made a mistake but this hindrance is only minor and overall I had a positive experience with
digitising.

Conclusion:

Summary Discussion :

- First of all, a new understanding of QGIS has been gained, and amazed by the power.
Apart from the difficulty of operating, QGIS can achieve much more functionality
than previously expected. Project 3 offered a great opportunity to practice and
cooperate in the field of mapping, it also prompts ability development for group 5,
such as digitizing, calculating, and also the ability to overcome the unexpected issues
(compatibility of software and system).
- It is also necessary to mention our experience of exploring the WMS connection.
Rather than repeating the power of the function, instead, I would like to emphasize
the new logic it brought to us. Using WMS induces us to explore the function in not
only QGIS but also external likelihood. I think it is much more meaningful than its
power.
- To improve the project, it is a good choice to highlight time using and logical
flexibility. Technique issues are unavoidable during a project but it would be rather
discouraged to waste too much time on it. Looking for help and the ability to find
alternatives is also essential in group work.