6th International Cloud Modeling Workshop (Hamburg, Germany. 11-16 July 2004)

Case 3: Heavy winter orographic precipitation (IMPROVE-2 case)

Organizers: Greg Thompson and Roy Rasmussen
NCAR - Research Applications Program, P.O. Box 3000, Boulder, CO USA 80307-3000

gthompsn@ucar.edu

1. Overview of Case 3

Case 3 involves a significant wintertime orographic precipitation event that occurred on 13-14 Dec 2001 that was extensively sampled during the IMPROVE-2 field project (Stoelinga et al. 2003). The IMPROVE (Improvement of Microphysical Parameterization Through Observational Verification Experiment) second field exercise took place in the Oregon Cascade mountain range and included airborne in situ measurements, rawinsondes, radiometers, wind profilers, raingauges, and the NCAR S-band dual-polarization Doppler radar (S-Pol). During this event, a significant cyclonic storm and vigorous upper-level cold-frontal rainband produced extensive and deep clouds and precipitation throughout the study area (Woods et al. 2004).
Initially, we ask modelers to perform two-dimensional simulations with the supplied terrain, initial, and boundary condition data (provided below in Appendix A) to facilitate comparisons between all researchers. Additionally, modelers may simulate the full three-dimensional event if desired but we request the supplied initial and boundary condition data (found in Appendix B) be used for comparisons to existing/other model results. We anticipate that both bulk and bin microphysical parameterizations will be applied to this case.

2. Synoptic and mesoscale aspects

Table 1: Graphics showing synoptic and mesoscale aspects of case.
	0000 UTC 13 Dec	1200 UTC 13 Dec	0000 UTC 14 Dec
250 mb	Winds	Winds	Winds
300 mb	Winds	Winds	Winds
500 mb	Winds Temp/RH	Winds Temp/RH	Winds Temp/RH
700 mb	Winds Temp/RH	Winds Temp/RH	Winds Temp/RH
850 mb	Winds Temp/RH	Winds Temp/RH	Winds Temp/RH
Surface	MSLP	MSLP	MSLP 925 mb Theta-E
Satellite	Infrared satellite image loop Visible satellite image loop
Radar	Reflectivity (0.5° tilt) image loop Reflectivity (1.5° tilt) image loop

Graphics provided here in Table 1 reveal the synoptic and mesoscale aspects of this case. Aircraft in situ and other relevant microscale data will be presented at the workshop.

3. Two-dimensional simulations

The map below shows the study area of IMPROVE-2 and specific locations pertinent to the case that are further described in Table 2. On the lower part of the map is a profile of terrain taken along the cross-section (A-B) shown on the map. The terrain was extracted from 30-second Digital Elevation Model data distributed by the USGS (shown by black line in figure below) and smoothed (1-2-4-2-1 pass filter) to create the input terrain data for the 2D simulations (red line in figure below).
Results from a preliminary 2D simulation using MM5 and the bulk microphysical parameterization discussed in Thompson et al. (2004) are provided in Appendix C. For this simulation, a 2D version of MM5 was configured with 344 horizontal points spaced approximately 1 km apart. In the vertical, 50 levels were unevenly spaced from 20 m at the surface to a maximum of 750 m from 9 to 16 km. The model was initialized horizontally homogeneous with the data provided in Appendix A that were derived from a rawinsonde launched near Creswell, OR (marked 'Sonde' on the map below). The winds were obtained by taking the component along the 238° azimuth which matches the AB cross-section. The number of cloud droplets was constant across the domain at 25 cm^-3 whereas measurements by the Convair580 ranged from approximately 15 to 30 cm^-3.

Table 2: Significant observational assets for the 13-14 Dec 2001 case and 2D model cross-section. Also refer to map below.
Map notation	Description	Lat (deg)	Lon (deg)	Elev (m)	observational platforms
"SLE"	Salem, OR	44.900	-123.000	59	NWS METARs, rawinsonde
"EUG"	Eugene, OR	44.133	-123.217	114	NWS METARs
"Newport"	Newport, OR	44.57	-124.03	48	NOAA-ETL wind profiler and NWS METARs (KONP)
"Coos Bay"	Coos Bay, OR	43.42	-124.25	4	NWS METARs (KOTH - North Bend)
"S-Pol"	near Sweet Home, OR	44.386	-122.855	475	S-band, dual polarization Doppler radar
"Sonde"	Creswell, OR	43.920	-123.030	122	UW mobile rawinsonde
"Pass"	Santiam Pass (US Hwy 20)	44.420	-121.850	1451	UW rainguage, snow crystal obs
"6K5"	Sisters, OR	44.305	-121.543	975	PNNL remote sensing lab
"mb"	McKenzie Bridge	44.1823	-122.0838	488	S-band vertically pointing radar and wind profiler
"ISS2"	Irish Bend, OR	44.3600	-123.2110	85	wind profiler
"ISS3"	Black Butte Ranch	44.3790	-121.6790	1027	UW rawinsonde, wind profiler
"fc"	Falls Creek, OR	44.3944	-122.3722	530	UW raingauge
"jj"	Jump-off Joe Mountain, OR	44.3867	-122.1673	1067	UW raingauge
"sj"	Santiam Junction, OR	44.4350	-121.9450	1143	UW raingauge, radiometer
"co"	Corbett State Park, OR	44.4194	-121.7917	1300	UW raingauge
"A"	Pacific Ocean	43.1676	-124.7196	0	start of cross-section
"B"	lee of Cascade range	44.7639	-121.0365	?	end of cross-section

Total time for the simulation should be at least 6 hours. Since the event was characterized by deep synoptic lift in association with an upper-level front as well as strong orographic forcing, we decided to initialize the model with a pre-determined amount of cloud ice and snow at upper levels. This was done to mimic the effects of the upper-level forcing to match the microphysical observations more closely. The equations used to initialize the vertical profile of cloud ice and snow are found in Table 3 below and we request all modelers use these data if possible.
We request the following quantities be output to enable intercomparison between all researchers. Example graphics can be found in Appendix C.

Half-hourly 2-D cross-sections (plotted as height vs. distance) of mixing ratios and number concentrations of microphysical species: cloud water, rain, cloud ice, snow, graupel, aggregates, etc. Mixing ratio plots should have units of g/kg while number concentration plots should have number/liter except cloud drop concentration which should be number/cm³. To make comparisons with others' results simpler, please use a maximum altitude of 10 km for the vertical axis.
Time-series plots of maximum mixing ratios and number concentrations for X-points greater than 160 and for the length of the simulation.
Total precipitation at the ground in mm in an accumulation versus distance plot. This plot should contain a line for each hour that represents the total accumulation thus far in the total simulation. To make comparisons with others' results simpler, please use a maximum of 30 mm for the vertical axis.
Anyone using detailed microphysics is asked to create size distribution plots for the series of model gridpoints:
(X=160, Z=1.55, 2.1, 2.8, 3.7, 4.5, 5.8 km)
(X=190, Z=1.55, 2.1, 2.8, 3.7, 4.5, 5.8 km)
(X=240, Z=1.55, 2.1, 2.8, 3.7, 4.5, 5.8 km)
(X=265, Z=1.55, 2.1, 2.8, 3.7, 4.5, 5.8 km)
(X=275, Z=1.55, 2.1, 2.8, 3.7, 4.5, 5.8 km).
Two sets of size distribution plots are requested. One should show particle diameter in mm using a linear scale on the x-axis and number concentration with units of m^-3 mm^-1 on the y-axis. The other should show diameter in microns using a logarithmic scale on the x-axis and number concentration with units of m^-3 microns^-1 on the y-axis. Furthermore, to distinguish between cloud drops and drizzle or rain, we suggest a threshold diameter of 40 microns.

4. Three-dimensional simulation

Researchers who desire to simulate the full event in three dimensions are encouraged to do so using the data provided in Appendix B). To be consistent with simulations already performed by Garvert et al. 2004 and Colle et al. 2004, we request all modelers start their simulation at 0000 UTC 13 Dec 2001 and perform a 36-hour simulation (ending 1200 UTC 14 Dec). Thus far, analysis of this event is concentrated on aircraft flights within two hours of 0000 UTC 14 Dec, but some further analysis has also been done of a later flight from 0500 to 0600 UTC 14 Dec.

5. Summary

Researchers who are tentatively planning to simulate this case include: Matt Garvert (Univ. Washington), Istvan Geresdi (Pec, Hungary), Bill Hall, Roy Rasmussen, Axel Seifert, Greg Thompson (NCAR), and Ruby Leung (PNL). We look forward to a productive workshop and superb interactions with all researchers.

Appendix A: 2D Initialization Data

Table 3: Input data for 2D simulations.
ASCII-format data	GIF image
Topography profile (344 points)	See map image above
1^st column input data Actual Skew-T/Log-P data from "Sonde"	Skew-T/Log-P plot of model initial data Actual Skew-T/Log-P plot from "Sonde"
Equations to initialize ice and snow	Plot of initial ice and snow

Included in Table 3 are links to ASCII-format data necessary for initialization and time-dependent boundary conditions. Only the model's first column of data is provided. Researchers may chose to use as many vertical levels as desired but should interpolate from the provided data to additional levels and remainder of grid points such that the initial state is horizontally homogeneous. Furthermore, the input boundary condition at the left edge should be constant while the right edge boundary condition should be open.
For convenient reference, there are also links to GIF-format images of the input data.

Appendix B: Initialization Data for 3-D case

Table 4: Input data for 3D simulations.
**NOTE:** *To download files, right click and chose menu item 'Save Link As...'*
Time	Surface obs	Rawinsonde obs	Gridded pressure-level analyses/forecasts CAUTION! See note in table caption.
0000 UTC 13 Dec	METARs (raw) (simplified) (decoded)	RAOBs (raw) RAOBs (decoded) KSLE image	00z analysis
0300 UTC 13 Dec	METARs (raw) (simplified) (decoded)	RAOBs (raw) RAOBs (decoded) KSLE image	3-h forecast
0600 UTC 13 Dec	METARs (raw) (simplified) (decoded)		6-h forecast
0900 UTC 13 Dec	METARs (raw) (simplified) (decoded)		9-h forecast
1200 UTC 13 Dec	METARs (raw) (simplified) (decoded)	RAOBs (raw) RAOBs (decoded) KSLE image	12z analysis
1500 UTC 13 Dec	METARs (raw) (simplified) (decoded)		3-h forecast
1800 UTC 13 Dec	METARs (raw) (simplified) (decoded)	RAOBs (raw) RAOBs (decoded)	6-h forecast
2100 UTC 13 Dec	METARs (raw) (simplified) (decoded)	RAOBs (raw) RAOBs (decoded) KSLE image	9-h forecast
0000 UTC 14 Dec	METARs (raw) (simplified) (decoded)	RAOBs (raw) RAOBs (decoded) KSLE image	12-h forecast
0300 UTC 14 Dec	METARs (raw) (simplified) (decoded)	RAOBs (raw) RAOBs (decoded) KSLE image	15-h forecast
0600 UTC 14 Dec	METARs (raw) (simplified) (decoded)	RAOBs (raw) RAOBs (decoded) KSLE image	18-h forecast
0900 UTC 14 Dec	METARs (raw) (simplified) (decoded)		21-h forecast
1200 UTC 14 Dec	METARs (raw) (simplified) (decoded)	RAOBs (raw) RAOBs (decoded) KSLE image	24-h forecast

Included in Table 4 are links to all data necessary to perform a 36-h simulation of the case. Surface and upper-air point observation data are included in their raw format as well as a self-explanatory decoded format (both in ASCII text). Gridded pressure-level data are provided from the NCEP-AVN model analyses/forecasts in GRIB-format files. These files are binary so note the message in the table caption about proper downloading.

Appendix C: Sample graphics (2-D simulation)

Table 5: Sample results from 2D simulation.
	Loop	0-h fcst	1-h fcst	2-h fcst	3-h fcst	4-h fcst
Temperature and Vertical velocity	T/w	T/w	T/w	T/w	T/w	T/w
Theta-e and RH with respect to ice	θ_e/RH_i	θ_e/RH_i	θ_e/RH_i	θ_e/RH_i	θ_e/RH_i	θ_e/RH_i
Cloud water and ice	q_c, q_i	q_i	q_c, q_i	q_c, q_i	q_c, q_i	q_c, q_i
Snow and rain	q_s, q_r	q_s	q_s, q_r	q_s, q_r	q_s, q_r	q_s, q_r
Graupel	q_g		q_g	q_g	q_g	q_g
All microphysics	q_c, N_i, q_s, q_r, q_g		q_c, N_i, q_s, q_r, q_g	q_c, N_i, q_s, q_r, q_g	q_c, N_i, q_s, q_r, q_g	q_c, N_i, q_s, q_r, q_g
Precip. Accum.
Domain-maximum mixing ratio time-series

Included in Table 5 are links to graphics generated from a preliminary 2D MM5 simulation. Please match these graphics as closely as possible to facilitate intercomparison among groups. Where possible, match specific colors and contour values.

References

http://improve.atmos.washington.edu

Colle, B.A., M.F. Garvert, J.B. Wolfe, and C.F. Mass, 2004: Microphysical budgets and sensitivity studies for the 13-14 December 2001 IMPROVE-2 event, J. Atmos. Sci., submitted.

Garvert, M.F., B.A. Colle, and C.F. Mass, 2004: Synoptic and mesoscale evolution of the 13-14 December 2001 IMPROVE-II storm system and comparison with a mesoscale model simulation, J. Atmos. Sci., submitted.

Garvert, M.F., C.P. Woods, B.A. Colle, C.F. Mass, P.V. Hobbs, and J.B. Wolfe, 2004: Comparisons of MM5 model simulations of clouds and precipitation with observations for the 13-14 December 2001 IMPROVE-2 event, J. Atmos. Sci., submitted.

Stoelinga et al., 2003: Improvement of microphysical parameterization through observational verification experiment, Bull. Amer. Meteor. Soc., 84, 1807-1826.

Thompson, G., R.M. Rasmussen, and K. Manning, 2004: Explicit forecasts of winter precipitation using an improved bulk microphysics scheme. Part I: Description and sensitivity analysis, Mon. Wea. Rev., 132, 519-542.

Woods, C.P., M.T. Stoelinga, J.D. Locatelli, and P.V. Hobbs, 2004: Cloud structures, microphysical processes and synergistic interaction between frontal and orographic forcing of precipitation during the December 13, 2001 IMPROVE-2 event over the Oregon Cascades, J. Atmos. Sci., submitted.

Questions regarding this case can be directed to Greg Thompson (gthompsn@ucar.edu).