search

UMEP-dev / UMEP

Urban Multi-scale Environmental Predictor
https://umep-docs.readthedocs.io/
61 stars 15 forks source link

Solar Energy of Building Envelope producing negative results #521

Closed est987 closed 1 year ago

est987 commented 1 year ago

Describe the bug I'm running the Solar Energy of Building Envelope processor, and for some odd reason I am receiving negative results. I tried with an older project's files and it worked fine. Not sure what is the problem here? Maybe something is wrong with my DSM?

To Reproduce Steps to reproduce the behavior:

  1. Run Solar Energy of Building Envelope on attached DSM, WALL ASPECT and WALL HEIGHT rasters: CRS:6991 UTC: 2 (see screenshot )

Desktop (please complete the following information):

Screenshots image image

Data See Attached. DSM_RADIATION_DATA.zip

Additional context I have been working on a similar work process for many similar projects and worked so far without a problem. As I said, couldn't reproduce problem with a different DSM file from another project.

biglimp commented 1 year ago

Maybe something wrong with the CRS and your current version of QGIS? Try ro reproject to a more common CRS such as UTM and try again. Also, The solar radiation values in your metfile looks too low. Units should be W/m2.

est987 commented 1 year ago

I rebuild the model with a more common CRS (2039) and was still receiving negative values. So I created a new the MET file only for the six hours I'm interested with updated data (2013-2022 avg), and I received a much more reasonable result.

image

The reason the radiation values in MET file look small, is because they're divided by six. My understanding is that they are hourly values, while only representing 10 min (1/6 hour). If I don't do the division, I receive results two orders of magnitude larger than my expected range (0-4 kWh). Hope I'm not misunderstanding something.

Thank you for your help!

BeitDeganKrinaAvgMetFile_DIVIDE-SIX_UTC2.txt

biglimp commented 1 year ago

Ok, now I see. Sorry for not noticing before. SEBE must be executed with hourly data. It is stated in the manual but we should make that more clear to avoid misstake.

est987 commented 1 year ago

Thanks! Do you think it is better to calculate the hourly average and rebuild the MET file with only one line per hour?

biglimp commented 1 year ago

Yes, one line per hour.