From: Cgrer 3D Model Discussion Group <rss@rssforward.net>
Subject: WRF settings for CMAQ
A lot of group members are using WRF to generate meteorology for CMAQ. Please note that nested Lambert Conformal Conic grids generated with the WRF Domain Wizard typically result in tangent grids, where True Latitude 1 = True Latitude 2. MCIP doesn't always handle this correctly, so the developers commented out the option and made it an error. Sang-Rin and I worked around this by editing ll2xy_lam.F to call the tangent version ll2xy_lam_tan, and it appears to have worked for his grid. However, try to set WRF grids where True Lat 1 != True Lat 2 to ensure compatibility, and always test 1 day of meteorology with MCIP before running the full case!
From the MCIP FAQ, a few more WRF settings to note for better CMAQ simulations, some requiring registry edits before compiling WRF:
What extra fields do I need to turn on in the "wrfout" (history) file with the WRF Registry?
MCIP can make use of the following fields if they are available in the WRF output (and these fields are NOT part of the WRF output by default in WRFv3.1): fractional land use (LANDUSEF), aerodynamic resistance (RA), stomatal resistance (RS), vegetation fraction in the Pleim-Xiu LSM (VEGF_PX), roughness length (ZNT), and inverse Monin-Obukhov length (RMOL).
Do I have to use the Pleim-Xiu land-surface model in MM5/WRF to run MCIP?
No. Some of the MCIP/CMAQ physics probably work better with the Pleim-Xiu (PX) LSM than other land-surface models, but changes have been made to improve the use of MCIP and CMAQ for other LSMs. One thing to note is that the mixing within the planetary boudnary layer (PBL) can be more consistent between MM5/WRF and CMAQ if the Asymmetric Convective Model version 2 (ACM2), which is commonly used with the PX LSM, is used in MM5/WRF.
Why is there a variable to specify the reference latitude for the WRF Lambert conformal simulations?
The notion of a "center latitude" (which is required in the I/O API used by CMAQ) is not available in the WRF output. There is additional flexibility in how one can define simulation domains in WRF beyond what was available in MM5. Just going on what's available in the WRF output headers, it is impossible to "reconstruct" an existing MM5 domain that covers the same area under certain circumstances. Several users have inquired about matching MM5 and WRF domains and getting the I/O API headers consistent. (There are multiple ways of using the fields in the I/O API headers to represent the same domains.) Thus, the option for users to hand-pick the center (or reference) latitude was created in MCIP.
[1]
___
Source: http://cgrer3dmodel.blogspot.com/2010/01/wrf-settings-for-cmaq.html#comment-form
[1] <https://blogger.googleusercontent.com/tracker/4326166576961287278-3770614510477712573?l=cgrer3dmodel.blogspot.com>
From the MCIP FAQ, a few more WRF settings to note for better CMAQ simulations, some requiring registry edits before compiling WRF:
What extra fields do I need to turn on in the "wrfout" (history) file with the WRF Registry?
MCIP can make use of the following fields if they are available in the WRF output (and these fields are NOT part of the WRF output by default in WRFv3.1): fractional land use (LANDUSEF), aerodynamic resistance (RA), stomatal resistance (RS), vegetation fraction in the Pleim-Xiu LSM (VEGF_PX), roughness length (ZNT), and inverse Monin-Obukhov length (RMOL).
Do I have to use the Pleim-Xiu land-surface model in MM5/WRF to run MCIP?
No. Some of the MCIP/CMAQ physics probably work better with the Pleim-Xiu (PX) LSM than other land-surface models, but changes have been made to improve the use of MCIP and CMAQ for other LSMs. One thing to note is that the mixing within the planetary boudnary layer (PBL) can be more consistent between MM5/WRF and CMAQ if the Asymmetric Convective Model version 2 (ACM2), which is commonly used with the PX LSM, is used in MM5/WRF.
Why is there a variable to specify the reference latitude for the WRF Lambert conformal simulations?
The notion of a "center latitude" (which is required in the I/O API used by CMAQ) is not available in the WRF output. There is additional flexibility in how one can define simulation domains in WRF beyond what was available in MM5. Just going on what's available in the WRF output headers, it is impossible to "reconstruct" an existing MM5 domain that covers the same area under certain circumstances. Several users have inquired about matching MM5 and WRF domains and getting the I/O API headers consistent. (There are multiple ways of using the fields in the I/O API headers to represent the same domains.) Thus, the option for users to hand-pick the center (or reference) latitude was created in MCIP.
[1]
___
Source: http://cgrer3dmodel.blogspot.com/2010/01/wrf-settings-for-cmaq.html#comment-form
[1] <https://blogger.googleusercontent.com/tracker/4326166576961287278-3770614510477712573?l=cgrer3dmodel.blogspot.com>
A lot of group members are using WRF to generate meteorology for CMAQ. Please note that nested Lambert Conformal Conic grids generated with the WRF Domain Wizard typically result in tangent grids, where True Latitude 1 = True Latitude 2. MCIP doesn't always handle this correctly, so the developers commented out the option and made it an error. Sang-Rin and I worked around this by editing ll2xy_lam.F to call the tangent version ll2xy_lam_tan, and it appears to have worked for his grid. However, try to set WRF grids where True Lat 1 != True Lat 2 to ensure compatibility, and always test 1 day of meteorology with MCIP before running the full case!
From the MCIP FAQ, a few more WRF settings to note for better CMAQ simulations, some requiring registry edits before compiling WRF:
What extra fields do I need to turn on in the "wrfout" (history) file with the WRF Registry?
MCIP can make use of the following fields if they are available in the WRF output (and these fields are NOT part of the WRF output by default in WRFv3.1): fractional land use (LANDUSEF), aerodynamic resistance (RA), stomatal resistance (RS), vegetation fraction in the Pleim-Xiu LSM (VEGF_PX), roughness length (ZNT), and inverse Monin-Obukhov length (RMOL).
Do I have to use the Pleim-Xiu land-surface model in MM5/WRF to run MCIP?
No. Some of the MCIP/CMAQ physics probably work better with the Pleim-Xiu (PX) LSM than other land-surface models, but changes have been made to improve the use of MCIP and CMAQ for other LSMs. One thing to note is that the mixing within the planetary boudnary layer (PBL) can be more consistent between MM5/WRF and CMAQ if the Asymmetric Convective Model version 2 (ACM2), which is commonly used with the PX LSM, is used in MM5/WRF.
Why is there a variable to specify the reference latitude for the WRF Lambert conformal simulations?
The notion of a "center latitude" (which is required in the I/O API used by CMAQ) is not available in the WRF output. There is additional flexibility in how one can define simulation domains in WRF beyond what was available in MM5. Just going on what's available in the WRF output headers, it is impossible to "reconstruct" an existing MM5 domain that covers the same area under certain circumstances. Several users have inquired about matching MM5 and WRF domains and getting the I/O API headers consistent. (There are multiple ways of using the fields in the I/O API headers to represent the same domains.) Thus, the option for users to hand-pick the center (or reference) latitude was created in MCIP.
Source: http://cgrer3dmodel.blogspot.com/2010/01/wrf-settings-for-cmaq.html#comment-form
From the MCIP FAQ, a few more WRF settings to note for better CMAQ simulations, some requiring registry edits before compiling WRF:
What extra fields do I need to turn on in the "wrfout" (history) file with the WRF Registry?
MCIP can make use of the following fields if they are available in the WRF output (and these fields are NOT part of the WRF output by default in WRFv3.1): fractional land use (LANDUSEF), aerodynamic resistance (RA), stomatal resistance (RS), vegetation fraction in the Pleim-Xiu LSM (VEGF_PX), roughness length (ZNT), and inverse Monin-Obukhov length (RMOL).
Do I have to use the Pleim-Xiu land-surface model in MM5/WRF to run MCIP?
No. Some of the MCIP/CMAQ physics probably work better with the Pleim-Xiu (PX) LSM than other land-surface models, but changes have been made to improve the use of MCIP and CMAQ for other LSMs. One thing to note is that the mixing within the planetary boudnary layer (PBL) can be more consistent between MM5/WRF and CMAQ if the Asymmetric Convective Model version 2 (ACM2), which is commonly used with the PX LSM, is used in MM5/WRF.
Why is there a variable to specify the reference latitude for the WRF Lambert conformal simulations?
The notion of a "center latitude" (which is required in the I/O API used by CMAQ) is not available in the WRF output. There is additional flexibility in how one can define simulation domains in WRF beyond what was available in MM5. Just going on what's available in the WRF output headers, it is impossible to "reconstruct" an existing MM5 domain that covers the same area under certain circumstances. Several users have inquired about matching MM5 and WRF domains and getting the I/O API headers consistent. (There are multiple ways of using the fields in the I/O API headers to represent the same domains.) Thus, the option for users to hand-pick the center (or reference) latitude was created in MCIP.
Source: http://cgrer3dmodel.blogspot.com/2010/01/wrf-settings-for-cmaq.html#comment-form
