SOLWEIG issue

[sunt05]

Original report by andymac1091 (Bitbucket: andymac1091, GitHub: andymac1091).

The original report had attachments: aspect.jpg, buildingDSM.jpg, buildingoutlines.jpg, CDSM.jpg, DEM.jpg, height.jpg, maxinputsincludingCDSM.jpg, minimalinputs.jpg, mystudyareaSVF.jpg

---

Hi

I have attempted to use my own data to model MRT. However the output i am getting is rather counter-intuitive suggesting that trees are warmer than buildings. I have made sure of the following:

1. I have the most update version of QGIS (2.18.16) and UMEP (1.2) installed from the repository
2. All data is projected (EPSG: 32750) and the same resolution (1m)
3. The land cover data (where used) does not have evergreen or deciduous trees

I have run my analysis with the minimum requirements in SOLWEIG (e.g. building and ground DSM, DEM, SVF, height and aspect) and also all fields (e.g. the above and vegetation DSM, UMEP landcover). I have not used any meteorological data and left all other values constant. Both versions produce strange results. The former areas surrounding buildings seem to be warmer and the latter trees are warmer also. I have also run this exact analysis (with your data) on both test datasets which provide reasonable results (e.g. cooler where there are trees).

Therefore i think there must be an issue with my datasets, but i'm not sure what. The ground and building DSM has values of buildings and then ground in masl. The DEM is ground in masl and the canopy dsm (CDSM) has relative heights (as per the instructions and test data).

The only other issue i can see is that shadows appear using the test data but not my own.

Do you have any suggestions as to what is wrong with my input data or am i doing something else incorrect

Thanks

Andy

Please see attached screenshots (lighter colours are higher temps)

[sunt05]

Original comment by Fredrik Lindberg (Bitbucket: fredrik_ucg, ).

---

The new version of UMEP produce a RunInfo-file. Could you please provide us with the text included in that file for the run that produce the stange results.

[sunt05]

Original comment by andymac1091 (Bitbucket: andymac1091, GitHub: andymac1091).

---

Sure, find below the one used with all variables as descrbied above. Thinking it might possibly be my DEM/DSM/CDSM i also averaged these over my land cover (removing any noise) classes for a subset, however these replicated my strange results (see image, also below is the runinfo file for this) --------------------Run Info for previous post analysis-------------------- This file provides run settings for the SOLWEIG run initiated at: Wed, 21 Feb 2018 14:25:01 Version: SOLWEIG v2015a

SURFACE DATA Digital surface model (DSM): E:/paper4/subsets/subi_clip/elevation/resample/buildingDSMR1.tif Model domain: rows = 388, columns = 697 Projected referece system: WGS_1984_UTM_Zone_50S Geographical coordinate system: WGS 84 Latitude: -31.9455757071 Longitude: 115.819768669 UTC: 1 Transmissivity of light through vegetation: 0.03 Digital vegetation canopy model (CDSM): E:\paper4\subsets\subi_clip\elevation\resample\CDSMR1null.tif Trunkzone estimated from CDSM Trunkzone as percent of canopy height: 0.25 Landcover scheme active. Parameters taken from: C:/Users/Andy/.qgis2/python/plugins\UMEP\SOLWEIG/landcoverclasses_2016a.txt Landcover grid: E:/paper4/subsets/subi_clip/elevation/resample/all/LandcoverR1new.tif

METEOROLOGICAL FORCING DATA Meteorological file not usedYear: 1997.0 Day of Year: 157.0 Hour: 12.0 Minute: 30.0 Ait temperature: 23.0 Relative humidity: 30.0 Global radiation: 810.0 Diffuse radiation: 92.5 Direct radiation: 895.0

HUMAN EXPOSURE PARAMETERS Absorption, shortwave radiation: 0.7 Absorption, longwave radiation: 0.95 Posture of human body: Standing ENVIRONMENTAL PARAMETERS Albedo of walls: 0.2 Albedo of ground (not used if land cover scheme is active): 0.15 Emissivity (walls): 0.9 Emissivity of ground (not used if land cover scheme is active): 0.95

ADDITIONAL SETTINGS Human considered as a cylinder

-------Run info for averaged DSM/DEM/CDSM-------

This file provides run settings for the SOLWEIG run initiated at: Wed, 21 Feb 2018 17:28:05 Version: SOLWEIG v2015a

SURFACE DATA Digital surface model (DSM): E:\paper4\subsets\subi_Testaverage\buildingDSMMR.tif Model domain: rows = 246, columns = 331 Projected referece system: WGS_1984_UTM_Zone_50S Geographical coordinate system: WGS 84 Latitude: -31.9442956278 Longitude: 115.819785033 UTC: 1 Transmissivity of light through vegetation: 0.03 Digital vegetation canopy model (CDSM): E:\paper4\subsets\subi_Testaverage\CDSMMRnull.tif Trunkzone estimated from CDSM Trunkzone as percent of canopy height: 0.25 Landcover scheme inactive

METEOROLOGICAL FORCING DATA Meteorological file not usedYear: 1997.0 Day of Year: 157.0 Hour: 12.0 Minute: 30.0 Ait temperature: 23.0 Relative humidity: 30.0 Global radiation: 810.0 Diffuse radiation: 92.5 Direct radiation: 895.0

HUMAN EXPOSURE PARAMETERS Absorption, shortwave radiation: 0.7 Absorption, longwave radiation: 0.95 Posture of human body: Standing ENVIRONMENTAL PARAMETERS Albedo of walls: 0.2 Albedo of ground (not used if land cover scheme is active): 0.15 Emissivity (walls): 0.9 Emissivity of ground (not used if land cover scheme is active): 0.95

ADDITIONAL SETTINGS Human considered as a cylinder

[sunt05]

Original comment by Fredrik Lindberg (Bitbucket: fredrik_ucg, ).

---

Ok, thanks. This tells me that your model is in Western Australia, correct? If so, you need to change UTC offset accordingly which now is 1 (=CET).

[sunt05]

Original comment by andymac1091 (Bitbucket: andymac1091, GitHub: andymac1091).

---

Yes the location is Perth, Western Australia. Ah yes, i had to resolve this issue for the time series of meteorological data i'm also using. I think i must have been on auto pilot. Looks much better now, i feel rather silly. Thanks for the solution

[sunt05]

Original changes by Fredrik Lindberg (Bitbucket: fredrik_ucg, ).

---

changed state from "new" to "invalid"