Calculation time when using different size maps but same # of nodes

[ediviu]

Hi,

I've been trying to calculate resource grids with different vector maps.
What I've noticed is that it takes much longer to calculate a resource grid with larger maps although they would have the same number of nodes.

What I do is the following:

1. I use a first map, 7 x 3.5 km. Then I select a grid with 25 m of resolution and a certain space. This configuration would calculate around 40.000 nodes and it could take about 2 hours.

2. I use another map of the same location, however this time its size is 14 x 7 km (double). Then I select a grid with 25 m of resolution again and THE SAME SPACE. This configuration AGAIN would calculate around 40.000 nodes. The very same 40.000 nodes!!!. However, in this case, the calculation could take 40 hours.

How's that possible? After all, I'm calculating the same amount of nodes, the only difference is that the base map I'm using in the second case is bigger. Am I doing something wrong? Why does it take longer to calculate the same?

Thank you!

[Duncan]

In order to perform a site assessment, WAsP checks the entire map. Bigger maps take longer. A complete site assessment is performed for each node, so if the number of nodes stays the same and the map data grows, the job will take longer.

For this reason, if you care about the calculation time, it's well worth constraining the vector map to be only as big as you need for your assessments. Don't underestimate the calculation time cost of including irrelevant map data in the vector map.

Three tricks can help.

1) Cut the map, so that it does not extend further than you need for your calculations. This extension depends on the height of calculation and the kind of terrain. Generally, roughness changes further away are more significant than elevation changes.

2) Use the map editor to cut a circular map, since the corner data are irrelevant

3) Use the map editor to thin the contour density further from your sites. For example, if you've got a map with 10m contour intervals, then you could make a local map from that extending 1km in all directions, and then make a larger map and thin the contours to 50m. Recombining these maps will give you a map which WAsP can use efficiently.