Unless you are going to use your georeferenced map as a simple background image, the next natural step is to digitize elements from it. You have already done so in the exercises about creating vector data in Lesson: Creating a New Vector Dataset , when you digitized the school fields. In this lesson, you are going to digitize the forest stands’ borders that appear in the map as green lines but instead of doing it using an aerial image, you will use your georeferenced map.
The goal for this lesson: Learn a technique to help the digitizing task, digitizing forest stands and finally adding the inventory data to them.
Open your map_digitizing.qgs project in QGIS, that you saved from the previous lesson.
Once you have scanned and georeferenced your map you could start to digitize directly by looking at the image as a guide. That would most likely be the way to go if the image you are going to digitize from is, for example, an aerial photograph.
If what you are using to digitize is a good map, as it is in our case, it is likely that the information is clearly displayed as lines with different colors for each type of element. Those colors can be relatively easy extracted as individual images using an image processing software like GIMP. Such separate images can be used to assist the digitizing, as you will see below.
The first step will be to use GIMP to obtain an image that contains only the forest stands, that is, all those greenish lines that you could see in the original scanned map:
Now you can select the pixels in the image that are making up the forest stands’ borders (the greenish pixels):
Once you are done with the selection you need to copy this selection as a new layer and then save it as separate image file:
You could do the same process with other elements in the image, for example extracting the black lines that represent roads or the brown ones that represent the terrain’ contour lines. But for us, the forest stands is enough.
As you did in the previous lesson, you need to georeference this new image to be able to use it with the rest of your data.
Note that you don’t need to digitize the ground control points any more because this image is basically the same image as the original map image, as far as the Georeferencer tool is concerned. Here are some things you should remember:
Check that the new raster is fitting nicely with the original map.
Having in mind the digitizing tools in QGIS, you might already be thinking that it would be helpful to snap to those green pixels while digitizing. That is precisely what you are going to do next create points from those pixels to use them later to help you follow the forest stands’ borders when digitizing, by using the snapping tools available in QGIS.
Zoom in and see what the polygons look like. You will get something like this:
Next one option to get points out of those polygons is to get their centroids:
Now you can remove the rautjarvi_green_polygon layer from the TOC.
Change the symbology of the centroids layer as:
It is not necessary to differentiate points from each other, you just need them to be there for the snapping tools to use them. You can use those points now to follow the original lines much easily than without them.
Now you are ready to start with the actual digitizing work. You would start by creating a vector file of polygon type, but for this exercise, there is a shapefile with part of the area of interest already digitized. You will just finish digitizing the half of the forest stands that are left between the main roads (wide pink lines) and the lake:
Change the new layer’s symbology so that it will be easier to see what polygons have already been digitized:
Now, if you remember past modules, we have to set up and activate the snapping options:
With these snapping settings, whenever you are digitizing and get close enough to one of the points in the centroids layer or any vertex of your digitized polygons, a pink cross will appear on the point that will be snapped to.
Finally, turn off the visibility of all the layers except forest_stands and rautjarvi_georef. Make sure that the map image has not transparency any more.
A couple of important things to note before you start digitizing:
Now you can start digitizing:
If you were not prompted for the polygon id when you finished digitizing it, go to Settings ‣ Options ‣ Digitizing and make sure that the Suppress attribute form pop-up after feature creation is not checked.
Your digitized polygon will look like this:
Now for the second polygon, pick up the stand number 358. Make sure that the Avoid int. is checked for the forest_stands layer. This option does not allow intersecting polygons at digitizing, so that if you digitize over an existing polygon, the new polygon will be trimmed to meet the border of the already existing polygons. You can use this characteristic to automatically obtain a common border.
The part of the polygon that was overlapping the existing polygon has been automatically trimmed out and you are left with a common border, as you intended it to be.
Now you have two forest stands ready. And a good idea on how to proceed. Continue digitizing on your own until you have digitized all the forest stands that are limited by the main road and the lake.
It might look like a lot of work, but you will soon get used to digitizing the forest stands. It should take you about 15 minutes.
During the digitizing you might need to edit or delete nodes, split or merge polygons. You learned about the necessary tools in Lesson: Feature Topology, now is probably a good moment to go read about them again.
Remember that having Enable topological editing activated, allows you to move nodes common to two polygons so that the common border is edited at the same time for both polygons.
Your result will look like this:
It is possible that the forest inventory data you have for you map is also written in paper. In that case, you would have to first write that data to a text file or a spreadsheet. For this exercise, the information from the inventory for 1994 (the same inventory as the map) is ready as a comma separated text (csv) file.
Open the rautjarvi_1994.csv file from the exercise_data\forestry directory in a text editor and note that the inventory data file has an attribute called ID that has the numbers of the forest stands. Those numbers are the same as the forest stands ids you have entered for your polygons and can be used to link the data from the text file to your vector file. You can see the metadata for this inventory data in the file rautjarvi_1994_legend.txt in the same folder.
To add the data from the .csv file:
The data from the text file should be now linked to your vector file. To see what has happened, open the attribute table for the forest_stands layer. You can see that all the attributes from the inventory data file are now linked to your digitized vector layer.
The data from the .csv file is just linked to your vector file. To make this link permanent, so that the data is actually recorded to the vector file you need to save the forest_stands layer as a new vector file. Close the attribute table and right click the forest_stands layer to save it as forest_stands_1994.shp.
Open your new forest_stands_1994.shp in your map if you did not added yet. Then open the attribute table. You notice that the names of the columns that you just added are no very useful. To solve this:
To finish gathering the information related to these forest stands, you might calculate the area and the perimeter of the stands. You calculated areas for polygons in Lesson: Supplementary Exercise. Go back to that lesson if you need to and calculate the areas for the forest stands, name the new attribute Area and make sure that the values calculated are in hectares.
Now your forest_stands_1994.shp layer is ready and packed with all the available information.
Save your project to keep the current map presentation in case you need to come back later to it.
It has taken a few clicks of the mouse but you now have your old inventory data in digital format and ready for use in QGIS.
You could start doing different analysis with your brand new dataset, but you might be more interested in performing analysis in a dataset more up to date. The topic of the next lesson will be the creation of forest stands using current aerial photos and the addition of some relevant information to your dataset.