How to create a risk map

[MonicaGallego]

Hello all,

I have been working with this and after to try it 1 millions of times, I could not do it. I will explain:

I have to do a risk map of deforestation, so I would need to combine: map of risks from distances to roads (100km of distance, and risk values between 1 and 32), map of risk from distances to rivers(50 km of distance and risk values between 1 and 16), past deforestation map (risks value between 1-32 as a function of the density of deforested areas) and protected areas (risk value= 2). How can I do it?

I know that I must build each map and then combine all of them, but how?

Steps to take:

1. To create the map of risk from distances to roads and rails: first, buffer with 100 km of distance. Then I need to divide these 100 kms into 32 risk classes (each class with 3.125 km wide and 32 classes best represented the continuous field of risk from roads and rail), with areas closest fo the road/railway given highest risk (=32) and areas furthest away given lowest risk (=1).

I guess this is done with a buffer, but what kind? Euclidean distances, multiple ring buffer, normal  buffer... how?

2. To create the map of risk from rivers: I have to do the same procedure as before but with a distance of 50 km and risks between 1-16.

3. Now I should combine both maps, in order to obtain the risk from the two maps. I have done it with the sum of the raster calculator, but the thing is that I only get the common areas of both, and the rest disappears. How can I do this?

4. After that, I have to mix this result with the protected areas, so that everything that falls in a protected area will be reclassified as low risk (2).How?

5. And finally, I have to combine the last map with the deforestation density layer. This layer is divided into 32 classes based in the density values �??i.e. group of highest density values were given highest risk value (=32), and the group of the lowest density values were given lowest risk value (=1)... How now? I did it with raster calculator, but the same thing happens again, because It just gives me the common areas.

So, how can I do the preceding steps? and the final map? Any idea?

Thank you very much,

Monica

[JodieParkes1]

Hi Monica,

Have you had any luck with this task? I am looking in to risk mapping at the moment and am trying to find the best way to get started.

Thanks,

Jodie

[SepheFox]

You didn't mention if you are using vector or raster data, but I will give a solution for vector data, because that is the most likely way that rivers and roads are digitized. You also didn't explain how you want your final values to be calculated. Adding all the risk values to get a possible result of 80, with protected areas having value 2 is my best guess. The other important clarification, is what feature are you trying to get risk values for? A larger analysis area polygon that contains all your other features perhaps?

1. Multiple ring buffer from your roads. Each ring should have the appropriate value from 1-32.

2. Multiple ring buffer from rivers. Each ring should have the appropriate value from 1-16.

3. Union of both buffer layers. Now you should have fields for both the river and road risk values in your result.

4. Use the identity tool; your input features is your analysis area polygon, your identity features is the union.

5. Identity of analysis area polygon with protected areas. Your input features is the analysis area polygon, your identity features is the protected areas.

6. Identity of analysis area polygon with deforestation density layer (you may have to convert this to polygons first). your input features is the analysis area polygon, your identity features is the deforestation density.

7. Add a field called risk. Calculate the field by adding the three fields from your roads, rivers, and deforestation risk together in the field calculator.

8. Select by attribute where the protected area FID <> -1. Calculate the risk field to 2.

Jodie Parkes​, if this raises more questions, let me know. This should be a fairly easy process as long as you have defined the problem correctly, and you have all the necessary data layers.