Replies: 1 comment 2 replies
-
Well this is indeed crazy long creating the vegetation Röckle zones while it should not. I think that the main reason of this long time was a "bug" (input vegetation geometries were multipolygons and not exploded thus making spatial indexes useless). Still not incredible though. Maybe we could merge geometries having nearly the same height (using the vertical resolution as reason to merge two geometries). In general I would say that there is a need to work on the vegetation since most of trees in cities are isolated and some recent works have shown that the URock current scheme for vegetation is not suitable for such trees (cf. our paper describing the URock model). |
Beta Was this translation helpful? Give feedback.
-
I thought it would be good to have this discussion here so that others can follow.
@j3r3m1 , me and @nilswallenberg are about to start a city-wide (Malmö) run using URock. We have derived building and vegetation with URockPrepare. Running with only buildings is very fast but adding vegetation slows down the modell considerably. We assume it is cos the complexity of vegetation vectors (see below). What would be your suggestion on how to speed things up, i.e. how to simplify the vegetation ploygons without loosing loo much detatil in the output. We are only interest in near-surface wind speed (1.5 magl).
Fredrik and Nils
Beta Was this translation helpful? Give feedback.
All reactions