Skip to main content

Full text of "DTIC ADA602942: The Mesoscale Predictability of Terrain Induced Flows"

See other formats


DISTRIBUTION STATEMENT A: Approved for public release; distribution is unlimited. 


The Mesoscale Predictability of Terrain Induced Flows 

Dale R. Durran 
University of Washington 
Department of Atmospheric Sciences 
Box 351640 

Seattle, WA 98195-1640 

phone: (206) 543-7440 fax: (206) 543-0308 email: durrand@atmos.washington.edu 

Grant Number: N00014-06-1-0827 
http://www.atmos.washington.edu/~durrand/ 


LONG-TERM GOALS 

To develop an understanding of the predictability of small-scale atmospheric circulations appearing in 
forecasts generated by state-of-the-art high-resolution mesoscale models. Using previously collected 
observations and archived simulations perfonned using the Navy’s COAMPS model, as well as new 
simulations, we focus on assessing the predictability of winds, mountain waves and clear air turbulence 
(CAT) in the lee of the Sierra Nevada. 

OBJECTIVES 

Specific questions addressed in our research include: 

1. How sensitive are downslope winds to atmospheric conditions upstream of the mountain barrier? 

2. When such sensitivity is not extreme, can forecast errors in downslope winds and mountain-wave 
structure be linked to large characteristic errors in the atmospheric conditions forecast to occur on 
the upstream side of the mountains? Can systematic improvements in COAMPS be identified to 
remove these errors? 

3. What do ensemble forecasts indicate about the sensitivity of downslope winds and mountain waves 
to the upstream conditions, and how can such forecasts be best used to predict these events? 


The answers to these questions are of direct benefit to operational forecasters using COAMPS to 
produce aviation and other forecasts over complex terrain. Although our focus is on the forecasting of 
terrain-induced mesoscale disturbances, our findings are likely to be relevant to the predictability of 
other mesoscale phenomena. 


1 


Report Documentation Page 

Form Approved 

OMB No. 0704-0188 

Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and 
maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, 
including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington 

VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it 
does not display a currently valid OMB control number. 

1. REPORT DATE 

30 SEP 2009 2 - REPORT TYPE 

3. DATES COVERED 

00-00-2009 to 00-00-2009 

4. TITLE AND SUBTITLE 

The Mesoscale Predictability of Terrain Induced Flows 

5a. CONTRACT NUMBER 

5b. GRANT NUMBER 

5c. PROGRAM ELEMENT NUMBER 

6. AUTHOR(S) 

5d. PROJECT NUMBER 

5e. TASK NUMBER 

5f. WORK UNIT NUMBER 

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 

University of Washington,Department of Atmospheric Sciences,Box 
351640,Seattle,WA,98195-1640 

8. PERFORMING ORGANIZATION 

REPORT NUMBER 

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 

10. SPONSOR/MONITOR'S ACRONYM(S) 

11. SPONSOR/MONITOR'S REPORT 
NUMBER(S) 

12. DISTRIBUTION/AVAILABILITY STATEMENT 

Approved for public release; distribution unlimited 

13. SUPPLEMENTARY NOTES 

14. ABSTRACT 

15. SUBJECT TERMS 

16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 

___ ABSTRACT 

18. NUMBER 19a. NAME OF 

OF PAGES RESPONSIBLE PERSON 

a. REPORT b. ABSTRACT c. THIS PAGE Same 3S 

unclassified unclassified unclassified Report (SAR) 

6 


Standard Form 298 (Rev. 8-98) 

Prescribed by ANSI Std Z39-18 





APPROACH 


The P.I., together with Dr. James Doyle of NRL, Monterey and graduate student P. Alex Reinecke, 
used the COAMPS model to conduct a series of 70-member ensemble simulations of high-wind 
events observed during the Terrain-Induced Rotors Experiment (T-REX). By examining the ensemble 
spread, we obtained an unprecedentedly complete description of the sensitivity of mountain waves, 
CAT and downslope to small variations in the initial conditions. 

WORK COMPLETED 

We completed the analysis of the sensitivity of mountain waves, CAT and downslope winds to small 
perturbations in the upstream conditions. We also demonstrated a surprising sensitivity of the 
simulated mountain waves to numerical resolution. The results from these efforts are detailed below. 

RESULTS 

This study represents one of the first attempts to systematically document the sensitivity of downslope- 
wind forecasts to initial conditions in a fully non-linear, three-dimensional NWP mesoscale model. An 
ensemble of 70 different initial conditions is generated for each of two prototypical downslope-wind 
events from the T-REX special observing period: IOP 6 (25-26 March, 2006) and IOP 13 (16-17 April, 
2006). Consistent with the available data, most of the simulations of IOP-6 show a large-amplitude 
mountain wave with upper-level tropospheric wave breaking and severe downslope winds. In contrast, 
wave breaking was not present for the most of the IOP-13 simulations, and consistent with the 
observations, the downslope winds while still significant, were weaker than in IOP-6. The strong 
winds in IOP-13 were generated by a layer of high static stability flowing beneath a mid- and upper- 
tropospheric layer of low stability. 

In both cases, the individual ensemble members were ranked according to the forecast intensity of the 
near surface winds in a region along the lee slope of the Sierras (the “wind speed metric”), and the 10 
strongest and 10 weakest ensemble members were grouped into two subsets. 

For the wave-breaking simulations (IOP-6), initial-condition errors grow rapidly leading to large 
variability of the downslope-wind forecast. Fig. la, shows the forecast wind speed metric for the 
members of the strong subset, which can be compared with the same data for the members of the weak 
subset in Fig. lb. The wind speeds in the strong and weak subsets diverge rapidly between forecast 
hours 3 and 6. The difference in the subset means of the wind speed metric at hour six represent the 
difference between a very severe 41m s' 1 downslope winds and a mild 13 m s' 1 wind event. 

The same information is provided for the event with layered static stability, but no wave-breaking (IOP 
13) in Fig. 2. In this case the subsets diverge over a period of roughly eight hours. The 26 m s' 1 winds 
in the strong subset represent a moderate downslope while the 4ms' 1 winds in the weak subset are a 
non-event. 


2 



50 



40 


7 30 

t/j 

£ 20 


4 6 8 

hours 


Figure 1: Wind speed metric as a function offorecast hour for each of the members of the (a) 
strong subset and (b) weak subset. The subset mean is given by the heavy curve. The subset mean 

forecast winds at hour 6 differ by 28 m s' 1 . 


50 



hours 


Fig. 2: As in Fig. 1, except for the ten members of the weak subset. The subset mean 
forecast winds at hour 12 differ by 26 m s' 1 . 


3 













The difference between the means of the strong and weak subsets for these two events is further 
illustrated in Fig. 3. A very large amplitude mountain wave is present in the strong-subset simulations 
for the wave-breaking case (IOP 6), with vertical velocities exceeding 14 ms' 1 . The wave is much 
weaker in the weak-member subset, with vertical velocities just reaching 8 m s' 1 . As suggested by the 
surface wind speeds plotted in Fig. 2, the difference between the response of the strong- and weak- 
member subsets in IOP 13 is the difference between a moderate and a very weak mountain wave. 



B B’B B' 


Fig. 3: Isentropes of potential temperature (black curves) and vertical velocity (contoured at 4 m s' 1 
intervals, with the zero contour omitted. The IOP 6 cases are in the top row, those for IOP 13 are on 
the bottom. The weak subset means are in the left column, the strong subset means are on the right 


Upstream soundings from the mean of the strong- and weak-member simulations were examined for 
both cases just one hour before the time of the maximum winds in the strong subset. In IOP 6, the 
wave-breaking case, the differences in the cross-barrier wind speed, potential temperature, and Brunt- 
Vaisala frequency for the two subsets was generally less than radiosonde observational errors. Such 
very small differences in the upstream conditions suggest that deterministic operational forecasts may 
have difficulty accurately predicting the strength of downslope winds and CAT associated with 
mountain-wave breaking. 

For cases with strong low-level static stability, like IOP 13, in which wave breaking did not play a 
major role, the predictability time-scale appears to be somewhat longer. Upstream soundings from one- 
hour prior to the time of maximum wind show clear differences between the mean profiles for the 
strong- and weak-member subsets. For example, a 2-km deep layer of strong static-stability is present 
directly above crest level for the strong members, whereas, the crest-level static-stability is 


4 








































considerably weaker for the weak members. Downslope winds that develop in cases like IOP 13 
appear to have some degree of predictability at 6-hour lead times, but not at lead times of 12 hours. 

Regardless of the mechanism responsible for strong downslope winds, the rapid growth of forecast 
uncertainty for the IOP-6 and IOP-13 events are considerably shorter than the optimistic view of 
mountain-wave predictability presented in Klemp and Lilly (1975). Although the predictability results 
for these two events may not generalize to all mountain-wave and downslope-wind events, they call in 
question the idea presented in Anthes et al. (1985) that the predictability time scales originally 
suggested for mesoscale motions by Lorenz (1969) are too pessimistic for terrain induced flows. The 
difficulties encountered by Nance and Colrnan (2000) trying to forecast downslope winds with a 
mesoscale model may also be due, at least in part, to the large initial-condition sensitivities we have 
documented. 

IMPACT/APPLICATIONS 

Future operational forecasts of downslope winds, CAT and other terrain induced mesoscale 
circulations will need to take into account the demonstrated uncertainty in deterministic forecasts by 
using ensemble techniques. 

TRANSITIONS 

Alex Reinecke completed his Ph.D and accepted at postdoc at NRL Monterey. While a postdoc at 
NRL, he was offered tenure-track positions as an assistant professor at McGill and the North Carolina 
State University. Alex accepted a third offer of a pennanent position at NRL, where he continues to 
work closely with navy researchers on ensemble forecasting and mesoscale predictability. 

RELATED PROJECTS 

None 

REFERENCES 

Anthes, et al., 1985: Predictability of mesoscale atmospheric motions. Adv. Geophysics, 28B, 159-202. 

Klemp, J. B. and D. K. Lilly, 1975: The dynamics of wave-induced downslope winds. J. Atmos. Sci., 
32, 320-339. 

Lorenz, 1969: The predictability of a flow which possesses many scales of motion. Te/lus, 21, 289- 
307. 

Nance, L. B. and B. R. Coleman, 2000: Evaluating the use of a nonlinear two-dimensional model in 
downslope windstorm forecasts. Wea. Forecasting, 15, 717-729. 

PUBLICATIONS 

Reinecke, P.A., and D. R. Durran, 2008: “The over-amplification of gravity waves in numerical 
solutions to flow over topography.” Mon. Wea. Rev., 137, 1533-1549. 


5 



Reinecke, P.A., and D. R. Durran, 2009: “Initial condition sensitivities and the predictability of 
downslope winds.” J. Atmos. Sci., in press. 

HONORS/AWARDS/PRIZES 

2007 Alan Berman Research Publication Award (NRL) with James Doyle for the publication “Rotor 
and sub-rotor dynamics in the lee of three-dimensional terrain.” 


6