investigation of different parameterizations of the mixed liquid-ice sub-grid statistical cloud...
Post on 14-Jan-2016
219 Views
Preview:
TRANSCRIPT
Investigation Of Different Parameterizations Of The Mixed Liquid-ice Sub-grid Statistical
Cloud Scheme Over Greece
E. Avgoustoglou (euri@hnms.gr)Hellenic National Meteorological Service
13th COSMO General Meeting, Servizio Meteorolologico, Rome 4th to 9th September 2011
PRESENTATION OUTLINE
Highlights of the sub-grid liquid-ice mixed scheme
Test cases tabulation and model set up
Implementation of the scheme
Results for Radiation temperatures, Cloud Cover,
T_2M,
Conclusions and Comments
Stratiform Cloud Fraction (R) (Sommeria, Deardorff, 1976)
ilc
ice
il
vapour
v
total
w
cp
cli
qqqqqqq
qTc
L
d
,
re temperatupotential water total,
liquid water
heat on condensati phase" mixed" , 1: iilic LrLrL
liwsw dGdqqqHR
)(
0 ,1
0 ,0)(
x
xxH
22
2
22/12 221
1exp
12
1
wwliliwliq
w
q
wlili
q
qqr
rrG
lilili www qqq wli qwliqr
Highlights of the sub-grid liquid-ice mixed scheme
factor icing , :c
ii q
qTr
21
2
1 QerfR
sw qq
Q
2122 2 swsw qqqq
A first order approximation for is assumed in reference to and by using Clausius-Clapeyron equation cloud fraction R is given by
sq sq
Sommeria και Deardorff (1977) further approximated R empirically :
1R0 ,6.1
12
1
QR
In analogy, a SubGgrid Statistical (SGS) cloud scheme is implemented to COSMO model (Raschendorfer) where the stratiform cloud cover is approximated by a two-parameter relation:
1R0 ,1
B
QAR
Parameter A refers to cloud cover at saturation and Β refers to critical value of saturation deficit. The default values of these parameters are set 0.5 and 4.0 respectively.
Test Cases Choice and Model set-upSeven cases are under investigation:
# Starting Date
1 Jan. 1 2005 12UTC (d050101_12)
2 Dec. 24 2007 12UTC (d071224_12)
3 Apr. 25 2006 12 UTC (d060425_12)
4 May 9 2008 12 UTC (d080509_12)
5 May 1 2009 12 UTC (d090501_12)
6 April 27 2011 12 UTC (d110427_12)
7 January 9 2011 12 UTC (d110109_12)
COSMO_4.6 (Modified by M.R) Horizontal grid 0.06250 (~7 Km)
273273 points 40 vertical levels time step: 30 sec GME (Analysis 0.50) 3 hs 48 hs IBM HPC Cluster 1600 (P4+)
m
Model Domain
Implementation of the Sub-grid Cloud Schemes and Test Tabulation
1. cosmo_4.6_rh: (icldm_rad = 4) is a run of COSMO_v4.6 with the default relative humidity scheme
2. R_rh: (icldm_rad = 4) is a run of COSMO_v4.6 with the default relative-humidity scheme and the modifications of the mixed-phase scheme. (It gave a slightly different answer than the default version of cosmo_4.6_rh and was used as reference against the tests of the mixed-phase statistical scheme denoted below. )
3. R_mix_0.5_2.0: (icldm_rad = 2, clc_diag=0.5 and q_crit=2.0) is a run of COSMO_v4.6 with the mixed-phase statistical scheme
4. R_mix_0.5_3.0: (icldm_rad = 2, clc_diag=0.5 and q_crit=3.0) is a run of COSMO_v4.6 with the mixed-phase statistical scheme
5. R_mix_0.5_4.0: (icldm_rad = 2, clc_diag=0.5 and q_crit=4.0) is a run of COSMO_v4.6 with the mixed-phase statistical scheme (This is the standard old run from previous COSMO year with the default values for clc_diag and q_crit and was included for reference purposes)
6. R_mix_0.4_5.0: (icldm_rad = 2, clc_diag=0.4 and q_crit=5.0) is run of COSMO_v4.6 with the mixed-phase statistical scheme
7. R_mix_0.4_4.0: (icldm_rad = 2, clc_diag=0.4 and q_crit=4.0) is run of COSMO_v4.6 with the mixed statistical scheme. (This test was performed only for one test case.)
8. ER_stat_rh_0.5_4.0: (clc_diag=0.5 and q_crit=4.0) is based on a modified COSMO 4.11 version that invokes statistical (not mixed!) scheme when there is no cloud ice at the grid point and the relative humidity scheme when there is cloud ice. (This is an old run related to a personal modification that was demonstrated in the presentation in Moscow.)
Considered Variables
CLCT: Total Cloud Cover
CLCH: High Cloud Cover
CLCM: Medium Cloud Cover
CLCL: Low Cloud Cover
CLC: Cloud Cover over station
T_2M: 2 meter Temperature
Cloud T: Artificial Satellite Images (MSG WV 6.2 microns)
! A larger set is available for further consideration and discussion with COSMO coleagues..
The results for the mixed-phase statistical scheme are presented for the “Spring Case” of April 27th 2011 12 UTC (mainly) and the “Winter Case” of January 9th 2011 12 UTC in reference to the default relative-humidity scheme, satellite pictures (MSG Infrared, Water vapor and Cloud Cover) as well as station observations.
Results
Tmin
6
7
8
9
10
11
12
13
14
15
16
17
18
19
606
614
619
622
624
627
628
632
641
642
643
648
650
654
655
665
667
671
672
675
682
684
685
687
699
706
710
715
716
718
719
723
726
732
734
741
742
744
746
749
750
754
757
759
760
765
766
768
774
R_rh
R_mix_0.5_4.0
R_mix_0.5_2.0
Observations
Deg C
Station #
28-April-2011 T2m_min
Minimum daily 2-meter Temperatures for 28th April 2011. The vast majority of values from the mixed- phase scheme runs (R_mix_0.5_4.0 and R_mix_0.5_2.0) are relatively closer to Observations than the corresponding values of the default relative humidity scheme of COSMO model (R_rh)
12
13
14
15
16
17
18
19
20
21
22
23
24
606
614
619
622
624
627
628
632
641
642
643
648
650
654
655
665
667
671
672
675
682
684
685
687
699
706
710
715
716
718
719
723
726
732
734
741
742
744
746
749
750
754
757
759
760
765
766
774
R_rh
R_mix_0.5_4.0
R_mix_0.5_2.0
Observ ations
28-April-2011 T2m_max
Station #
Maximum daily 2-meter Temperatures for 28th April 2011. The values from the default relative humidity scheme of COSMO model (R_rh) are relatively closer to Observations than the corresponding values of the mixed- phase scheme runs (R_mix_0.5_4.0 and R_mix_0.5_2.0) but the situation is more balanced than the corresponding minimum daily 2-meter Temperatures .
Deg C
28 April 2011 03 UTC
Meteosat/Synesat R_rh R_mix_0.5_4.0
R_mix_0.5_2.0 R_mix_0.4_5.0 ER_stat_rh_0.5_4.0
Deg C
Comparison of cloud radiation temperatures (CRTs) of artificial satellite images at the water-vapor Channel at 6.2 microns against Meteosat satellite image (upper right) manipulated with Synesat software available at HNMS. The general trend (especially in the encircled areas) is that model CRTs for the mixed-phase scheme (R_mix_0.5_4.0, R_mix_0.5_2.0 and R_mix_0.4_5.0) are overall lower to those of the default relative humidity scheme (R_rh) in reference to Meteosat satellite picture (Meteosat/synesat). However, it looks that this difference is mainly due to the treatment of cloud ice by the mixed-phase scheme as it can be infered by a “control” test were the relative-humidity and the statistical scheme are mixed (ER_stat_rh_0.5_4.0).
28 April 2011 02:45 UTC %
Meteosat/Synesat R_rh R_mix_0.5_4.0
Low Cloud Cover (LCC)
Comparison of Low Cloud Cover (LCC) against the corresponding MSG satellite image (Meteosat/Synesat) manipulated with Synesat software available at HNMS. The general trend (especially in the encircled areas) is that the model LCC for the mixed-phase scheme (R_mix_0.5_4.0) show an overall preponderance to the LCC of the default relative-humidity scheme (R_rh) in reference to Meteosat satellite image. Another interesting feature is that in reference to the satellite picture, the mixed-phase scheme provides a better tendency to resolve the cloudiness in contrast to the cloud structure of the relative humidity scheme which has the tendency to remain more compact. The feature is similar to Medium Cloud Cover.
28 April 2011 02:45 UTC
Meteosat/Synesat R_rh R_mix_0.5_4.0
%Medium Cloud Cover (MCC)
Comparison of Medium Cloud Cover (MCC) against the corresponding MSG satellite image (Meteosat/Synesat) manipulated with Synesat software available at HNMS. As in LCC, the general trend (especially in the enclosed areas) is that the model MCC for the mixed-phase scheme (R_mix_0.5_4.0) show an overall preponderance over the MCC of the default relative-humidity scheme (R_rh) in reference to the satellite image. Again, similarly to LCC, we note that in reference to the satellite picture, the mixed-phase scheme provides a better tendency to resolve the cloudiness, while the cloud structure of the relative humidity scheme has the tendency to remain more compact.
28 April 2010 02:45 UTC
Meteosat/Synesat R_rh R_mix_0.5_4.0
%High Cloud Cover (HCC)
Comparison of High Cloud Cover (HCC) against the corresponding MSG satellite image (Meteosat/Synesat) manipulated with Synesat software available at HNMS. In reference to MSG picture, both cloud schemes overestimate HCC, especially over the West/South-West parts, addressing the issue of proper accounting of cloud-ice content. In contrast to LCC and MCC and in reference to the satellite picture, the relative-humidity scheme provides a better tendency to resolve the cloudiness, while the cloud structure of the mixed-phase scheme has the tendency to remain more compact.
28 April 2010 02:45 UTC
Meteosat/Synesat R_rh R_mix_0.5_4.0
%
Comparison of Total Cloud Cover (TCC) against the corresponding MSG satellite image (Meteosat/Synesat) manipulated with Synesat software available at HNMS. Both cloud schemes are in very good agreement with each other as well as with MSG picture. This feature demonstrates that the relative differences in Low, Middle and High cloud-cover layers between the two schemes are converging towards very similar Total Cloud Cover.
Total Cloud Cover (TCC)
%R
_rh
R_m
ix_0
.5_4
.0
TCC
TCC
HCC
HCC
MCC
MCC
LCC
LCC
Cloud Cover for 11 January 2011 (12 + 06:00 UTC)
Model Cloud Cover (TCC, HCC, MCC, LCC) against the corresponding “common” MSG satellite image (MSG 3.9 microns), as it is available in the University of Dundee Database and for a winter case due to unavailability of Synesat data. In reference to MSG image, we may see that TCC is overestimated for both schemes but the mixed-phase scheme performs better over Eastern parts. This preponderance is mainly due to the differences in LCC and MCC between the two schemes. Regarding HCC both cloud schemes are in very good agreement with each other but in contrast to the previous “spring” test case, HCC is overestimated for both schemes as is shown in MSG picture and this has a strong effect against the agreement with the satellite picture.
MSG 3.9 microns
%
R_rh R_mix_0.5_4.0
Low Cloud Cover
MSG 3.9 microns
Low Cloud Cover
For low clouds in particular, and for the “winter” test case (January 9th-10th 2011), it was noticed that the bulk of low cloud cover for the COSMO model default relative humidity (R_rh) scheme interchanges position with the mixed statistical scheme (R_mix_0.5_4.0).
CONCLUSIONS-COMMENTSDue to the nature of the project, satellite data are of prime importance regarding the evaluation of the investigated sub-grid cloud schemes. There was an effort to make use of MSG data as they can be manipulated at HNMS through the SYNESAT system available locally.Due to limitations regarding older data, a new recent “Spring” test case was chosen, for which satellite data were at the time available in the local SYNESAT station.SYNESAT provided a nice flexibility regarding the comparison of satellite data and model output and quite a few relative differences of the sub-grid cloud schemes under consideration were highlighted. It looks that, in reference to satellite data, the mixed-phase scheme pushes the model towards colder radiation temperatures than the default relative-humidity scheme. The overall cloud cover is decently compared with the satellite figures for both schemes.Regarding medium and low clouds, it looks that the mixed-phase scheme preponderates in areas with relatively low cloudiness and the relative humidity scheme preponderates in more cloudy areas.It should be noted that regarding minimum 2m Temperatures the mixed-phase scheme performs better comparing with observations while the situation for the maximum 2m Temperatures is more balanced between the two schemes.For low clouds in particular, by carefully choosing another “winter” recent test case it was detected that the bulk of low cloud cover for the COSMO model default relative humidity scheme interchanges position with the mixed-phase statistical scheme. On the basis of the current as well as past works, the overall effect regarding the operational implementation of the sub-grid mixed-phase statistical cloud scheme to COSMO model over the wider geographical area of Greece may considered overall neutral. However, the overall slight improvement of minimum daily 2m-temperature as well the middle and low cloudiness over marine areas should be considered decent assets. The implementation of the SGS mixed-phase scheme as an option to a later COSMO model version will be of value for both operational and research purposes in the COSMO community and is recommended although a systematic investigation will be of value to fully evaluate the mixed-phase scheme.
top related