Search Results
NMC level. III data werein general available twice a day at 0000 and 1200 GMT,and were used for comparison together with the level IIIdata (map) of ANMRC (Australian Numerical Meteorology Research Centre) in Melbourne.564 JOURNAl. OF THE ATMOSPHERIC SCIENCES VO~.UME339 SEP 197400 GMTSURFACEFro. lc. FIo. 1. Examples of data maps for 00 GMT=t=3 h, 9 September 1'974. (a) 200 mb map includes the data in the layer
NMC level. III data werein general available twice a day at 0000 and 1200 GMT,and were used for comparison together with the level IIIdata (map) of ANMRC (Australian Numerical Meteorology Research Centre) in Melbourne.564 JOURNAl. OF THE ATMOSPHERIC SCIENCES VO~.UME339 SEP 197400 GMTSURFACEFro. lc. FIo. 1. Examples of data maps for 00 GMT=t=3 h, 9 September 1'974. (a) 200 mb map includes the data in the layer
. So D e is the size averaged throughout the cloud layer, which at 2111 UTC was about 3.5 km thick. The satellite retrievals of Ou et al. (1995) show that values of D e about 60 μ m are present around Coffeyville at 2111 UTC. However, according to the satellite data map, the Coffeyville hub was located within a region of high horizontal gradient of D e . To compare size information from different sources, it should be expressed in the same terms as was done for comparisons shown in Fig. 2
. So D e is the size averaged throughout the cloud layer, which at 2111 UTC was about 3.5 km thick. The satellite retrievals of Ou et al. (1995) show that values of D e about 60 μ m are present around Coffeyville at 2111 UTC. However, according to the satellite data map, the Coffeyville hub was located within a region of high horizontal gradient of D e . To compare size information from different sources, it should be expressed in the same terms as was done for comparisons shown in Fig. 2
, and convective accelera-tion are computed fromq=m2k.VX-+fVmwhere m is the map-scale factor, k a vertical unit vector,V is the horizontal grad operator in a pressure surface,f the Coriolis parameter, v the horizontal vector velocitywith t,he components u and 0. The numerical value ofthe constant c is computed in Appendix 1.The following relation was used as horizontal boundaryconditionwp=o=wp,*m=o.From this and the continuity equation it follows that6=0, so that m-l; may be defined by a stream
, and convective accelera-tion are computed fromq=m2k.VX-+fVmwhere m is the map-scale factor, k a vertical unit vector,V is the horizontal grad operator in a pressure surface,f the Coriolis parameter, v the horizontal vector velocitywith t,he components u and 0. The numerical value ofthe constant c is computed in Appendix 1.The following relation was used as horizontal boundaryconditionwp=o=wp,*m=o.From this and the continuity equation it follows that6=0, so that m-l; may be defined by a stream
. The fittedparameters are consistent with independent estimates of the errors in the wind stress analysis. The calibratederror model is used in a Kalman filtering scheme to generate monthly sea level height anomaly maps for thetropical Pacific. The filtered maps, i. e., those which result from data assimilation, exhibit fine structure that isabsent from the unfiltered model output, even in regions removed from the data insertion points. Error estimates,an important byproduct of the scheme, suggest
. The fittedparameters are consistent with independent estimates of the errors in the wind stress analysis. The calibratederror model is used in a Kalman filtering scheme to generate monthly sea level height anomaly maps for thetropical Pacific. The filtered maps, i. e., those which result from data assimilation, exhibit fine structure that isabsent from the unfiltered model output, even in regions removed from the data insertion points. Error estimates,an important byproduct of the scheme, suggest
available ( NOS 1999 ). The COFS nowcast serves the initial conditions for a daily 24-h forecast and the next day's assimilation cycle. The purpose of this paper is to present the results of comparisons between COFS temperature predictions with and without SST data assimilation. COFS SST predictions were evaluated for a 3-month period from 1 January to 31 March 1998, and subsurface predictions were assessed for February of 1998. A description of the version of the coastal ocean circulation model used
available ( NOS 1999 ). The COFS nowcast serves the initial conditions for a daily 24-h forecast and the next day's assimilation cycle. The purpose of this paper is to present the results of comparisons between COFS temperature predictions with and without SST data assimilation. COFS SST predictions were evaluated for a 3-month period from 1 January to 31 March 1998, and subsurface predictions were assessed for February of 1998. A description of the version of the coastal ocean circulation model used
for stratiform instability and convectively coupled waves. J. Atmos. Sci , 58 , 1567 – 1584 . Majda , A. J. , B. Khouider , G. N. Kiladis , K. H. Straub , and M. G. Shefter , 2004 : A model for convectively coupled tropical waves: Nonlinearity, rotation, and comparison with observations. J. Atmos. Sci , 61 , 2188 – 2205 . Mapes , B. E. , 1998 : The large-scale part of tropical mesoscale convective system circulations: A linear vertical spectral band model. J. Meteor
for stratiform instability and convectively coupled waves. J. Atmos. Sci , 58 , 1567 – 1584 . Majda , A. J. , B. Khouider , G. N. Kiladis , K. H. Straub , and M. G. Shefter , 2004 : A model for convectively coupled tropical waves: Nonlinearity, rotation, and comparison with observations. J. Atmos. Sci , 61 , 2188 – 2205 . Mapes , B. E. , 1998 : The large-scale part of tropical mesoscale convective system circulations: A linear vertical spectral band model. J. Meteor
[+, VV] fields, where +=(g/jo)+, V`=( m/d)2V2, and J= (m/2d)']. The relative vorticity and vor-ticity advection are then written, respectively, as follows:AAAandwhere jo is j at +45O N., d is the NWP grid distance of381 km. at 6=60 N., and m=(l+sin 60)(1+sin +)",the map factor.3. CASE SELECTION AND REANALYSIS PROCEDURESince our objective is the improvement of 500-mb.initial analyses in data-sparse areas, the followingprocedure of case selection has been used. The errorcharts (forecast minus verifying
[+, VV] fields, where +=(g/jo)+, V`=( m/d)2V2, and J= (m/2d)']. The relative vorticity and vor-ticity advection are then written, respectively, as follows:AAAandwhere jo is j at +45O N., d is the NWP grid distance of381 km. at 6=60 N., and m=(l+sin 60)(1+sin +)",the map factor.3. CASE SELECTION AND REANALYSIS PROCEDURESince our objective is the improvement of 500-mb.initial analyses in data-sparse areas, the followingprocedure of case selection has been used. The errorcharts (forecast minus verifying
. Verification Radio soundings launched a few kilometers away from the downstream endpoint of the automobile's traverse during the field phase of MAP allows one to compare slantwise reduction with vertical reduction and estimate the error from partly neglecting horizontal changes in the vertical temperature profile. This comparison was particularly exacting for two reasons. The comparison location is farthest—both vertically (∼800 m) and horizontally (∼30 km)—away from the reduction level. And often there
. Verification Radio soundings launched a few kilometers away from the downstream endpoint of the automobile's traverse during the field phase of MAP allows one to compare slantwise reduction with vertical reduction and estimate the error from partly neglecting horizontal changes in the vertical temperature profile. This comparison was particularly exacting for two reasons. The comparison location is farthest—both vertically (∼800 m) and horizontally (∼30 km)—away from the reduction level. And often there
thelines on a transparent overlay, shown in fig. 2, wasplaced on a point where the smoothed value was tobe computed. The line b-d was always oriented northsouth. The smoothed value obtained was equal to(a+l.+c+d+e)/S, and the distance (f) was equal to25 deg lat. A comparison of a smoothed map with anunsmoothed map showed about what would be ex= 1 Xsec-I, 8 X 23 Xsurface of non-divergence might be located from thetwelve cross sections. It resulted that a surface of leastdivergence could be located at
thelines on a transparent overlay, shown in fig. 2, wasplaced on a point where the smoothed value was tobe computed. The line b-d was always oriented northsouth. The smoothed value obtained was equal to(a+l.+c+d+e)/S, and the distance (f) was equal to25 deg lat. A comparison of a smoothed map with anunsmoothed map showed about what would be ex= 1 Xsec-I, 8 X 23 Xsurface of non-divergence might be located from thetwelve cross sections. It resulted that a surface of leastdivergence could be located at
change ofabout 0.1C in 12 hr at these latitudes. As may be seen later, thisis smaller than R2At. For the purpose of this study, equation (12)is sufficient.2 74 JOURNAL OF METEOROLOGY VoLixI?: 16FIG. 1. Positions of an air particle and of an isobaric surface at two different times.a finite interval of timei.e., AAt, may be evaluatedfrom data on isobaric maps.In fig. 1, surfaces p(tl) and P(t2) represent thehypothetical positions of an isobaric surface at timestl and t2, respectively. B(x2, y
change ofabout 0.1C in 12 hr at these latitudes. As may be seen later, thisis smaller than R2At. For the purpose of this study, equation (12)is sufficient.2 74 JOURNAL OF METEOROLOGY VoLixI?: 16FIG. 1. Positions of an air particle and of an isobaric surface at two different times.a finite interval of timei.e., AAt, may be evaluatedfrom data on isobaric maps.In fig. 1, surfaces p(tl) and P(t2) represent thehypothetical positions of an isobaric surface at timestl and t2, respectively. B(x2, y
