rbell@rsmas_miami_edu
New Member
Hi, I am a post-doc working at the University of Miami with Prof. Kirtman and Dr. Min. I am working on using the output from NMME to run a wave model (WaveWatchIII) and investigate seasonal prediction of Ocean waves which is why I require the 10m U and V components.Firstly, Apologies for asking yet another question on 10m U and V output in CESM/CAM/CCSM4(https://bb.cgd.ucar.edu/how-caculate-wind-speed-10m-using-u-and-v-ccsm4-0; https://bb.cgd.ucar.edu/surface-winds-reversed-sign; https://bb.cgd.ucar.edu/creating-near-surface-zonal-and-meridional-wind-output-variables; https://bb.cgd.ucar.edu/northward-eastward-near-surface-wind-speed-variables; https://bb.cgd.ucar.edu/regarding-roughness-length-z0; https://bb.cgd.ucar.edu/surface-wind-speed-cmip5-simulation-output; https://bb.cgd.ucar.edu/any-documentation-out-there-detailed-explanation-cam-output-fields; https://bb.cgd.ucar.edu/node/1002195; https://bb.cgd.ucar.edu/node/1000871) I have two questions. 1) I would like to calulate to calculate the wind components at 10m using data in the suite of hindcasts we have already run. 2) Going forward I would like add 10m wind components in Min's ongoing operational seasonal forecasting using CCSM4.1) We have the data for TAUX, TAUY and U, V (wind on sigma levels). I have attached a screen show of the calculation for U-10 (10m U wind component). Unfortunately we don't have Z0MG (momentum roughness length over ground, in m). NCEP has surface roughness data (ftp://ftp.cdc.noaa.gov/Datasets/ncep.reanalysis.dailyavgs/surface_gauss/sfcr.sfc.gauss.${year}.nc) but only over the ocean (and low spatial resloution). ERA-Interim has surface roughness length but it only varies in space and not time. ERA-I has the charnock parameter though which can be used to caluclate surface roughness length everwhere (http://glossary.ametsoc.org/wiki/Charnock's_relation). I have decided to use surface roughness length from ERA-Interim. To smooth out the storms I have created a climatology of surface roughness (daily). I use this to calculateion U-10. However, in doing this I get some very large values for U-10 (and V-10) (larger than the values at U_bot). I've attahced screen shots of U_bot, TAUX, Z0 and U-10 for one time. I'll now look at putting flags in my code that if U-10 is larger than U_bot scale it down by some amount. I have now noticed that large wind components occurs when U_bot is a different sign to TAUX. i.e. the calcuation for U-10 becomes > than U_bot, using the absolute value of TAUX. If anybody has looked at the before any help would be appreciated.2) Unfortunately I am not running (have not/may not run) CCSM4 and may be limited in my understanding of the output diagnostics. I am hoping to gather as much information as possible and point Min in the right direction for testing. It seems you can output U10 (total wind speed) by adding 'U10' to the namelist. What was the logic behind only outputting total wind speed and not the components? they are output for U,V,TAUX and TAUY and total wind speed can be calculated easily given the components. I assume it is possible in the code which outputs U10 to cut it one line short and output the vectors (I am assuming the last line of the code is U10=sqrt(U-10^2 + V-10^2)). Has anyone done this before? If I wanted to output Z0MG (momentum roughness length over ground, in m) do I (Min) simply add this to the namelist?Happy to clarify further.Thanks,Ray