Dear WRF Community,
I am running WRF to improve hurricane wind and pressure fields, which I plan to use as forcing for other models that require regular lat-lon data. My target domain is the Gulf of Mexico, predominantly covering low latitudes (12-35N- parent domain) and extending wider east-west than north-south. I initially tried the Mercator projection but would prefer to avoid re-projecting each time, especially since I plan to conduct several sensitivity tests on the wrf-outputs for the other models, and I also use MATLAB extensively.
I noticed in the WRF manual that regular lat-lon can be used for regional domains by rotating the projection so that the area of interest is near the equator of the projection. Based on the manual, I set up my WPS namelist as follows:
map_proj = 'lat-lon',
ref_lat = 24.00,
ref_lon = -88.00,
pole_lat = 66.
pole_lon = 180
stand_lon = 88.0,
Q1. a. Is my namelist configuration correct for achieving a rotated lat-lon grid (I'm interested in regular lat-lon)?
b. The WRF output metadata shows "MAP_PROJ_CHAR = 'Cylindrical Equidistant'" and the output itself looks okay except for some trimming towards low lat as shown below. Should I be concerned that this isn't labeled as "rotated"? Is the output still in regular lat-lon even when using a rotated setup (which seems the case here), or is the term "Cylindrical Equidistant" simply referring to the overall lat-lon structure?
c. would I need to rotate the 10-meter wind components (u10, v10) when processing the outputs for my other models?
Q2: Would Lambert be a better choice for my domain, or should Mercator suffice (if I decide to return to Mercator)?
Q3: For both projections in Q2, would I need to rotate the 10-meter wind components (u10, v10) when processing them for my other models?
Thank you for your time and guidance.
I am running WRF to improve hurricane wind and pressure fields, which I plan to use as forcing for other models that require regular lat-lon data. My target domain is the Gulf of Mexico, predominantly covering low latitudes (12-35N- parent domain) and extending wider east-west than north-south. I initially tried the Mercator projection but would prefer to avoid re-projecting each time, especially since I plan to conduct several sensitivity tests on the wrf-outputs for the other models, and I also use MATLAB extensively.
I noticed in the WRF manual that regular lat-lon can be used for regional domains by rotating the projection so that the area of interest is near the equator of the projection. Based on the manual, I set up my WPS namelist as follows:
map_proj = 'lat-lon',
ref_lat = 24.00,
ref_lon = -88.00,
pole_lat = 66.
pole_lon = 180
stand_lon = 88.0,
Q1. a. Is my namelist configuration correct for achieving a rotated lat-lon grid (I'm interested in regular lat-lon)?
b. The WRF output metadata shows "MAP_PROJ_CHAR = 'Cylindrical Equidistant'" and the output itself looks okay except for some trimming towards low lat as shown below. Should I be concerned that this isn't labeled as "rotated"? Is the output still in regular lat-lon even when using a rotated setup (which seems the case here), or is the term "Cylindrical Equidistant" simply referring to the overall lat-lon structure?
c. would I need to rotate the 10-meter wind components (u10, v10) when processing the outputs for my other models?
Q2: Would Lambert be a better choice for my domain, or should Mercator suffice (if I decide to return to Mercator)?
Q3: For both projections in Q2, would I need to rotate the 10-meter wind components (u10, v10) when processing them for my other models?
Thank you for your time and guidance.