1. The Use of Conceptual Models in the Forecast Process for Frontal Precipitation Events Steven Koch North Carolina State University COMAP 99 9 August 1999

2. Technological advances in observing, numerical modeling, and computational capabilities have led to increasing reliance by forecasters upon direct interpretation of model forecast results. Do you agree with this statement often heard?……….. “ Why show me approximations to the governing equations when today’s numerical weather prediction models make no such approximations, contain all the important physics, and have sufficient resolution? The model fields are more useful in operations than any theory or conceptual model based on a set of assumptions. ” Why can this attitude be dangerous? What is the need for frontal conceptual models?

3. This statement implies faulty understanding of NWP model capabilities. Meteorological analysis, understanding, and prediction cannot proceed successfully independent of theory and conceptualization. visualize J. Bjerknes: It is necessary to visualize the atmosphere in three dimensions to grasp its complexity in a way that no ordinary plane image can ever show (such as from a surface chart alone).

4. What is the need for frontal conceptual models? model hangs together Sutcliffe (1942): It is necessary to “mould the whole model around the incomplete scaffolding of observations using theory, experience, flair, scientific judgment or even aesthetic sense to provide something which hangs together.” inadequate The simple Norwegian frontal model is inadequate at explaining the great diversity of frontal structures and precipitation distributions in the mid-latitude cyclone. need Forecasters need conceptual models to make sense of cloud and precipitation patterns observed and forecast.

5. Outline Frontal Conceptual Models:  Anafronts and Katafronts  Conveyor Belt Framework  Anafront with Low-level Jet  Narrow Cold Front Rainband  Katafront with dry intrusion: Split Fronts and Cold Fronts Aloft Narrow Cold Frontal Rainband — 28 Dec 88 (Maryland) Palm Sunday Tornado Outbreak — 27 Mar 94 (Alabama) Split Front Cases — 19 Dec 95 and 22 Jan 99 (SE U.S.). These cases form the labs today and tomorrow.

6. Relative Flow Patterns Anafront Katafront Bergeron (1934): Front-relative flow visualizations: Anafront — rearward ascent (post-frontal precip) Katafront — forward ascent (pre-frontal precip) Carlson (1980) and Browning (1985) recast this concept in the framework of relative isentropic-flow analysis to explain satellite cloud imagery (conveyor belts). mutual and Accordingly, it is the mutual consideration of the front- relative flow (U-C) in the plane of frontal motion and the moisture field that determines the type of frontal system and which can explain the distribution of clouds and precipitation.

7. Frontal Potpourri Split front Split front: a katafront with mid-level intrusion of low  e air detached from and ahead of surface cold front that overruns the warm conveyor belt. Cold Front Aloft Cold Front Aloft: a katafront with mid-level intrusion of low  e ahead of a surface pressure trough or dryline (Hobbs et al. 1990; 1996) Note: the conveyor belt paradigm does not include any theoretical basis for its predictions, unlike the Sawyer-Eliassen equation in semi-geostrophic theory.

8. Split fronts and Cold Fronts Aloft are important mechanisms for prefrontal bands of strong convection, including above Cold Air Damming. Anafront with strong southerly low-level jet may generate an intense but shallow Narrow Cold Front Rainband (NCFR) at surface cold front and post- frontal slantwise precipitation, often in the form of Wide Cold Frontal Rainbands (WCFR). The NCFR is not well appreciated in the aviation community and is poorly forecast. The slantwise ascent can result in a long-duration precipitation event with embedded convection. Importance to Character of Precipitation

9. NCFR Event of 28 December 1988 (Koch and Kocin 1991 Meteor. and Atmos. Phys.) Unforecast Unforecast gale-force winds with >100 mm h -1 rainrates for only 15 min, followed by snow NCFR event: 35 m s -1 LLJ, “precip cores and gaps”, though with lee slope diminution of slantwise precipitation Strong subsidence associated with tropopause fold  isallobaric center on windward slopes  explosive ageostrophic frontogenesis + frontal merger  development of NCFR and “cool pool” (gravity current)  NCFR sustained by vorticity balance (between “cool pool” solenoid and LLJ vertical shear)

10. Forecasting Suggestions from this case possibility of NCFR development Carefully monitor possibility of NCFR development from front-relative flow and multispectral satellite imagery. Be alert for frontal mergers. Frontogenesis isallobaric convergence Frontogenesis and isallobaric convergence should be used as predictive tools in the forecast office. cross sections Construct cross sections normal to fronts of mixing ratio, ageostrophic circulation, isentropes. tropopause folds. Analyze IPV to search for tropopause folds. Relate to vertical motions and to surface isallobaric field. Calculate Calculate cross-front vorticity balance.

11. 11 supercell storms, 32 tornadoes, 42 deaths in a narrow corridor from northern AL to western NC but ahead of All 11 storms contained within one MCS composed of 2 lines parallel to, but ahead of, slow-moving cold front (two other MCSs occurred right along the front) No quasi-geostrophic forcing No quasi-geostrophic forcing in outbreak region:  No Q-vector convergence or QG low-level frontogenesis  Upper-level jet transverse circulation was not coupled to the low-level frontal circulation Palm Sunday Tornado Outbreak: 27 March 1994 Koch et al. 1998 (Mon. Wea. Rev.)

12. Palm Sunday Tornado Outbreak: 27 March 1994 Koch et al. 1997 (Mon. Wea. Rev. in press) Then why was there strong ascent in all the models in northern Alabama? ageostrophic  Strong ageostrophic frontogenetical forcing for the two frontal bands of severe, but non-tornadic storms (i.e, anafront NCFR)  Supercells asymmetric inertial instability.  Supercells in lines 100-200 km ahead of front explained by theory of asymmetric inertial instability. Inertial instability arose due to a mid-level jetlet  Jetlet was spawned by earlier convection in Texas due to a geostrophic adjustment process.  Circulation transverse to this jetlet occurring in presence of strong convective instability (CAPE > 3000) spawned the supercells.

13. Forecasting Suggestions from this case may not even couple Circulation transverse to upper-level jet is not the Holy Grail; furthermore, it may not even couple with the circulation transverse to the low-level front. the single Apply the relevant conceptual model to help interpret vertical cross sections normal to fronts. Remember: frontogenesis can only explain a single rainband directly along a front (not ahead of it!) Precip forecast conceptual model Precip forecast should take into account nature of the coupling and the correct conceptual model. Examine Examine MesoEta model, profiler, 88D VWP for jetlets