I reported on this a few months ago. The code is wrong and itwas confirmed by the wrf group who is responsible for the script. Please check the ncl archives.
Regards, Andrea
Andrea Hahmann
Copenhagen, Denmark
On 17/03/2012, at 03.36, "李嘉鹏" <lijpbasin@126.com<mailto:lijpbasin@126.com>> wrote:
Hi,
The code in "$NCARG_ROOT/lib/ncarg/nclscripts/wrf/WRFUserARW.ncl" lines from 770 to 781 might be incorrect!
770 cone = 1.
771 if( map_projection .eq. 1) then ; Lambert Conformal mapping
772 if( (fabs(true_lat1 - true_lat2) .gt. 0.1) .and. \
773 (fabs(true_lat2 - 90. ) .gt. 0.1) ) then
774 cone = 10^(cos(true_lat1*radians_per_degree)) \
775 -10^(cos(true_lat2*radians_per_degree))
776 & nbsp; cone = cone/(10^(tan(45. -fabs(true_lat1/2.)*radians_per_degree)) - \
777 10^(tan(45. -fabs(true_lat2/2.)*radians_per_degree)) )
778 else
779 cone = sin(fabs(true_lat1)*radians_per_degree)
780 end if
781 end if
lines from 774 to 776 should be
774 cone = log(cos(true_lat1*radians_per_degree)) \
775 -log(cos(true_lat2*radians_per_degree))
776 cone = cone/( log(tan(( 45. -fabs(true_lat1/2.))*radians_per_degree)) - \
777 log(tan((45. -fabs(true_lat2/2.))*radians_per_degree)) )
according to the following web pages:
http://mathworld.wolfram.com/LambertConformalConicProjection.html
http://en.wikipedia.org/wiki/Lambert_conformal_conic_projection
The code which calculates cone factor in ARWpost V3.1's source file
ARWpost/src/module_map_utils.f90
is also provided (lines 634 to 642)
634 IF (ABS(truelat1-truelat2) .GT. 0.1) THEN
635
636 ! Compute cone fac tor following:
637 cone=(ALOG(COS(truelat1*rad_per_deg))-ALOG(COS(truelat2*rad_per_deg))) / &
638 (ALOG(TAN((90.-ABS(truelat1))*rad_per_deg*0.5 ))- &
639 ALOG(TAN((90.-ABS(truelat2))*rad_per_deg*0.5 )) )
640 ELSE
641 cone = SIN(ABS(truelat1)*rad_per_deg )
642 ENDIF
NCL version: 6.0.0
my system:
Linux balance 3.2.9-1-ARCH #1 SMP PREEMPT Thu Mar 1 09:31:13 CET 2012 x86_64 AMD Phenom(tm) 9150e Quad-Core Processor AuthenticAMD GNU/Linux
--
LiJiapeng (lijpbasin@126.com<mailto:lijpbasin@126.com>)
<名片.vcf>
<ATT00001.c>
_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Sat Mar 17 00:19:29 2012
This archive was generated by hypermail 2.1.8 : Tue Mar 20 2012 - 15:27:15 MDT