Kunena Site Syndication http://hirlam.org Wed, 19 Jun 2013 19:26:21 +0000 Kunena 1.0.11 http://hirlam.org/components/com_kunena/template/default_ex/images/english/emoticons/rss.gif http://hirlam.org Kunena Site Syndication http://hirlam.org/index.php?option=com_kunena&Itemid=160&func=view&catid=2&id=1085#1097 For the future learning: in all subroutines, which call something below them, like src/arp/phys_dmn/apl_arome.F90 calling e.g. aro_ground_param.F90 there is defined an interface block with #include aro_ground_param.intfb.h . As far as I understand, this include has been written there manually but it is used by makeup to create the block itself automatically and then use in linking? Now, sometimes such block #includes are missing for some routines, still the system mainly works? Is it possible to end up to problems because of the missing definitions? Can this have any relation to the problems discussed above? Wed, 19 Jun 2013 16:45:56 +0000 http://hirlam.org/index.php?option=com_kunena&Itemid=160&func=view&catid=2&id=1093#1093 Dear all, I would like to do a sensitivity run with decreased surface roughness. I found two roughness file in the constant files directory, $HM_CLDATA (/project/hirlam/harmonie_climate/), one in SURFACE_G and the other one in SURFACE_L. I assume G is global and L is for higher resolution, but without being able to inspect them I do not know anything. The constant files are binaries. Do you know how I can inspect them, i.e. make them readable or visualise them? Has anyone made amendments to constant files? All the best, Nína Tue, 18 Jun 2013 15:32:54 +0000 http://hirlam.org/index.php?option=com_kunena&Itemid=160&func=view&catid=2&id=1077#1092 Hi Enda, thanks for putting this in. Pages like this are really nice to have for reference when trying to make improvements. Cheers, Nína Tue, 18 Jun 2013 15:25:48 +0000 http://hirlam.org/index.php?option=com_kunena&Itemid=160&func=view&catid=2&id=921#1084 Dear Dagrun, The issues described above have been hard-coded into the IFS-radiation routines that are used in HARMONIE - presumably by the Meso-NH group. So as long as the AROME-IFS radiation is used, you cannot get the direct SW irradiance out. Best regards, Kristian. Fri, 14 Jun 2013 12:54:52 +0000 http://hirlam.org/index.php?option=com_kunena&Itemid=160&func=view&catid=2&id=1062#1082 Laura, other, We are looking for a better solution in cy38 than to just sleep a little longer. However, I've changed the cy37h1.2 setup in the mean time. Ulf Wed, 12 Jun 2013 08:28:17 +0000 http://hirlam.org/index.php?option=com_kunena&Itemid=160&func=view&catid=2&id=913#1079 Just testing the defaults (both climate datasets & code) of cy38 & ecoclimap2 Some roughly 30-40 % of the Curonian Lagoon neat Lithuanian coast appears to be land rather than sea (which it actually is, not only the LSM itself but the surface params also indicating that pretty well (not attached)) with ECO-2. ECO-1 was more or less OK with this place. http://www.hirlam.org/images/fbfiles/images/arorain30a_6.png Tue, 28 May 2013 11:12:26 +0000 http://hirlam.org/index.php?option=com_kunena&Itemid=160&func=view&catid=19&id=1078#1078 In the attached image the noise in a 40 hour forecast of hourly global horizontal SW irradiance (GHI) is shown. The forecast time is at night, where the GHI should be zero. The noise level is up 5 W/m² and most likely arises from the lack of precision in the GRIB data and the very large values of the accumulated flux - typically 10's of millions of J/m². The question then is, why do we store accumulated fluxes in units of [J/m²]? Why not just store hourly accumulated fluxes in units of [J/m²/hour] or hourly average fluxes in units of [Wh/m²/h]? Doing this is likely to remove the noise problem. http://hirlam.org/images/fbfiles/images/glorad_noise_example.jpg Fri, 24 May 2013 09:48:58 +0000 http://hirlam.org/index.php?option=com_kunena&Itemid=160&func=view&catid=2&id=1073#1076 Hi Trygve, Thanks for this answer, I think it explains everything! Yes, I'm using AROME (and cy37h1.1), and yes, it was just the first cycle I tested. The NEINI=0 report was in the Forecast log-file, since I got no separate log-file for DFI. I'll try again over 2 or more cycles... Thanks!! -Enda Wed, 22 May 2013 12:04:46 +0000