GEOG 7010 - Literature Reviewscottkehler.powweb.com/portfoliodocs/ElevatedConvectionReview.pdf · Elevated convection was defined by Colman (1990a) as thunderstorms that occur above

GEOG 7010 - Literature Review

Written By: Scott Kehler – 7687711

December 9, 2015

1. Introduction

1.1 Elevated Convection Background

Rainfall has long been known to exhibit a nocturnal maximum across the Great Plains of

the United States during the warm season (e.g. Kincer, 1916). This nocturnal maximum has

been attributed to the frequent occurrence of Mesoscale Convective Systems (MCS) during

this season (Maddox, 1980). The development of convective storms after dark has long

represented a challenging research question. These storms usually form above a stable

boundary layer, which differs considerably from the typical model of daytime convection

whereby parcels rise from the surface due to solar heating. Storms that form above a stable

boundary layer have been termed “elevated convection” by Colman (1990a). More recently,

Corfidi et al. (2008) have proposed a “continuum of convection” whereby convection ranges

from purely surface-based to purely elevated. Elevated convection is by now well embedded

within the broader literature about convection (e.g. Colman 1990a, Colman 1990b, Horgan et

al., 2007; Marsham et al., 2011; Moore et al., 2003; Rochette and Moore, 1996; Trier et al.,

2006). However, there remains a distinct lack of research regarding the fundamental

processes that lead to elevated convection (many past studies are climatology-based). This

can be partly attributed to the difficulty in acquiring data that is representative of elevated

convection environments. Since elevated convection occurs above the boundary layer, the

driving processes are occurring in a data poor part of our observing network. Rawinsondes or

specialized remote sensing instruments (e.g. AERI, radiometer, lidar, etc.) are required to

observe the troposphere above the surface. Unfortunately, our current upper-air observing

network relies mainly on rawinsondes which are of low spatial and temporal resolution, and

satellites, which have higher temporal resolution, but are not well suited to observe the fine

details of the lower troposphere. In recognition of these issues, the Plains Elevated

Convection at Night (PECAN) field project was conducted between June 1 and July 15, 2015

to improve our understanding of elevated convection. This field project had the following

scientific objectives (Geerts et al., 2013):

1. Initiation and early evolution of elevated convection

2. MCS internal structure and microphysics

3. Bores and wave-like features

4. Storm and MCS-scale NWP

Given the difficulty in acquiring observations of elevated convection, Numerical Weather

Prediction Models (NWP) can be an important tool to study elevated convection (and the

atmosphere in general). These models provide us with data about environments that is much

higher resolution than our observing network. They also give us the opportunity to conduct

sensitivity studies, whereby certain variables can be altered to assess their impact on a given

phenomena. Various models from a range of agencies are available to conduct research, but

one such model, the Weather Research and Forecasting Model (WRF), is perhaps the best

example of a research-oriented modeling system. The WRF is available in the public

domains for use and modification. The flexibility afforded by the WRF has made it a popular

choice among atmospheric researchers worldwide.

This paper summarizes our current knowledge of elevated convection (background to

PECAN objective 1) and discusses the WRF in the context of simulating elevated convection

(background to PECAN objective 4).

2. Elevated Convection – Our Current Understanding

Elevated convection was defined by Colman (1990a) as thunderstorms that occur above

frontal surfaces and are isolated from surface diabatic effects. Horgan et al. (2007) and

Corfidi et al. (2008) more generally defined elevated convection as occurring above a near-

surface stable layer. This near-surface stable layer could be a frontal or nocturnal inversion.

The author prefers the Corfidi et al. (2008)/Horgan et al. (2007) definition since elevated

convection does not necessarily occur above a frontal surface (e.g. convection above a

nocturnal inversion). Corfidi et al. (2008) also proposes a continuum of convection, where

convection ranges from purely surface-based to purely elevated, but also between convection

that is driven by latent heat release and that associated with thermals rising through the LFC.

This continuum recognizes that our current understanding of convection is not sufficient to

always definitively classify convection as one type or the other. Figure 1 shows examples of

two different types of elevated convection environments. Figure 1 (a) shows an example of

elevated instability above a frontal inversion. This profile was taken at Chanhassen, MN on

May 10, 2011 at 1200 UTC, approximately 200 km north of warm front. It has a frontal

inversion below 900 mb and a weak easterly surface flow (not shown). Figure 2 (b) shows a

profile taken at Hesston, KS on July 9, 2015 at 0400 UTC in a region with no apparent

frontal features and a light south-easterly surface flow (not shown). There is no pronounced

nocturnal inversion in the profile, rather it shows the nocturnal boundary layer at 11 pm local

time (the boundary layer is not well mixed and relatively stable). Both panels in Figure 1

show conditional instability, despite the fact that parcels originating from the surface have no

useable instability. Figure 1 (a) has Most Unstable Convective Available Potential Energy

(MUCAPE) of 5500 𝐽 𝑘𝑔−1 (parcel originating near ~900 mb) and Figure 1 (b) has

MUCAPE of 170 𝑘𝑔−1 (parcel originating near ~650 mb). Despite these MUCAPE values,

SBCAPE in Figure 1(a) is 0 𝐽 𝑘𝑔−1 and only 70 𝐽 𝑘𝑔−1 in Figure 1(b). Even though the

surface parcel in Figure 1(b) exhibits conditional instability, it has surface-based convective

inhibition of -170 𝐽 𝑘𝑔−1, suggesting the instability will not be released (whereas the

MUCAPE parcel has almost no convective inhibition). Another notable characteristic of the

profile in Figure 1 (a) is the steep lapse rates of ~9 °𝐶 𝑘𝑚−1 around 600 mb. Like surface-

based convection, elevated storms can benefit from the presence of these steep mid-level

lapse rates, also known as an Elevated Mixed Layer (EML). The presence of an EML would

tend to increase a storm’s updraft velocity, thereby potentially supporting larger hail. The

profiles shown in Figure 1 are just examples of what an elevated convection environment

could look like. Just like surface-based conditional instability, the presence of elevated

instability does not mean a convective storm will develop, it merely indicates potential for

development. Perhaps the most difficult question in this regard is determining when elevated

conditional instability is present, and the likelihood of it being released, given our sparse

upper-air observation network. Further to this, the trigger for elevated convection can also be

more difficult to ascertain, given that it too may not be reflected at the surface. The following

parts of this section discuss possible mechanisms for the development of elevated

convection.

2.1 Role of the Low-Level Jet

The low-level jet is known to exhibit a nocturnal maximum over the Great Plains of the

United States (Bonner, 1968). Low-level jets occur in other parts of North America, and

indeed the world, but the frequency of strong low-level jets is unique to the Great Plains

(Bonner, 1968).

Figure 1: (a) Rawinsonde observation at Chanhassen, MN showing a frontal inversion on

May 10, 2011 at 1200 UTC. (b) The nocturnal boundary layer at Hesston, KS on July 9, 2015

at 0400 UTC. The parcel trace (dashed) shows MUCAPE in both panels.

A

B

The low-level jet occurs at ~800 m above ground level (AGL), thus influencing the lower

troposphere (Bonner, 1968). This low-level wind maximum is believed to be a contributing

factor to the nocturnal rainfall maximum on the Great Plains due to its role in generating

strong warm air advection (WAA) and rapid moisture transport (Maddox, 1983). In fact,

Maddox (1983) suggests the lift generated by this strong low-level WAA is larger than the

lift generated by differential positive vorticity advection (PVA).

2.2 Fronts

Fronts are characterized by the following attributes (Lackmann, 2011):

Enhanced horizontal contrasts of temperature and/or moisture; moisture gradients

alone may suffice if we accept the “air mass boundary” definition of fronts

A relative minimum of pressure (trough) and maximum of cyclonic vorticity along

the front

Strong vertical wind shear, and a horizontal wind shift consistent with a pressure

trough and cyclonic vorticity

Large static stability within the frontal zone

Ascending air, clouds, and precipitation near the front (depending on moisture

availability and other factors); and

Greatest intensity near the surface, weakening with height

Frontogenesis is the process by which a front strengthens; conversely frontolysis is the

process by which a front weakens. Frontogenesis is described by Equation 1 (for the

direction perpendicular to the front):

(1) 𝐹 =𝜕𝜃

𝜕𝑥(

𝜕𝑢

𝜕𝑦) +

𝜕𝜃

𝜕𝑦(

𝜕𝑣

𝜕𝑦) +

𝜕𝜃

𝜕𝑝(

𝜕𝜔

𝜕𝑦) −

𝜕

𝜕𝑦(

𝑑𝜃

𝑑𝑡)

Where F is the frontogenesis function with four terms (from left to right): shearing,

confluence, tilting, and diabatic effects. Shearing, confluence, and tilting refer to changes in

the temperature gradient brought about by the u, v, and w components, respectively, of the

wind. The diabatic term refers to other effects that can affect the temperature gradient, such

as differences in insolation across the front.

Elevated convection is typically associated with warm fronts due to warm air over-riding

a cooler air mass. Warm fronts are characterized by an advancing warm air mass and

retreating cold air mass. These fronts advance by thermal advection and/or turbulent mixing.

Frontal advance can occur more quickly during the day as thermal mixing is enhanced and

static stability is reduced (Lackmann, 2011). The front itself has strong static stability,

countering the movement by turbulent mixing. The depth of warm fronts is not constant, thus

making their identification challenging at times. Since shallow warm fronts are not likely to

be well reflected in surface analyses, upper air soundings are often required for identification.

2.3 Elevated Convection Arising from Frontal Overrunning

Studies have shown a connection between low-level jets over-running surface fronts and the

development of Mesoscale Convective Systems (MCS) (e.g. Augustine and Caracena, 1994;

Moore et al., 2003; Trier and Parsons, 1993). Augustine and Caracena (1994) proposed a

physical model to predict the location of nocturnal MCS development. Their model used the

location of a surface geostrophic wind maximum along with the location of a surface front

and 850 mb frontogenesis to predict the location of a MCS. They propose that a mature MCS

is likely to be found directly poleward of a surface geostrophic wind maximum on the cool

side of a stationary front – assuming there is frontogenesis at 850 mb. Trier and Parsons

(1993) observed a rapid increase in CAPE due to transport of moisture up a frontal surface by

the low-level jet. They note that this can result in the rapid development of convection north

of a surface front as warm, moist air ascents the frontal surface. In the same study, Trier and

Parsons (1993) documented a case where the low-level jet rapidly transported high

equivalent potential temperature (𝜃𝑒) over a frontal surface. Rawinsonde observations at

Pratt, Kansas (PTT) measured lower tropospheric 𝜃𝑒 increasing from ~330 K to in excess of

350 K over a 3 h period as a result of moisture transport by the low-level jet. Using this

rawinsonde data along with surrounding sites, they produced a cross-sectional analysis

(Figure 2) showing a poleward sloping frontal surface being over-run by this low-level jet. A

MCS propagated north of the surface front, taking advantage of these favourable conditions.

Figure 2: Vertical cross section oriented from 346° (left) to 164° (right). Potential

temperature (𝜃)is subjectively analyzed in 4-K increments (solid lines), with selected 2-K

increments (dashed) added to help better delineate frontal structure. Horizontal winds are

plotted with standard meteorological conventions. From Trier and Parsons (1993).

2.4 Hazards from Elevated Convection

Elevated convection presents all the hazards of surface-based convection: hail, wind,

heavy rainfall, and tornadoes. However, it is well known that some phenomena are

considerably less likely to result from elevated convection – namely severe surface winds

and tornadoes. The general reasoning for this is the presence of a near-surface stable layer

(frontal/nocturnal inversion), which is not a feature of surface-based convection. However,

that explanation does not apply to all cases, as there are multiple documented cases in the

literature where severe surface winds and/or tornadoes have occurred even with a near-

surface stable layer (e.g. Colby and Walker, 2007; Goss and Thompson, 2006). Bryan and

Weisman (2006) used data from the Bow Echo and MCV Experiment (BAMEX) to study the

environmental conditions and mechanisms that produced severe surface winds from elevated

convection. They found that severe surface winds began to occur in a simulation of a squall

line once a surface cold pool developed. Based on this finding, they conclude that the surface

cold pool may be an important factor in determining if severe surface winds will occur in a

given system. However, they also note that further simulations will be required to test this

finding. Colby and Walker (2007) studied tornadoes resulting from elevated convection in

Iowa on May 21, 2004. They found that 8 of the 21 tornadoes on this day were from elevated

thunderstorms and that 6 of these 8 tornadoes occurred within an hour of each other.

However, while they are able to show the meteorological conditions that led to these

tornadoes, they conclude by stating that the mechanism allowing tornadoes to form from

elevated thunderstorms remains unknown. Goss and Thompson (2006) documented a case of

severe surface winds of up to 78 kt produced by an elevated supercell. They propose that the

environment in their case was favourable for what they term “overshooting downdrafts”

which can penetrate through a near-surface stable layer. Downdraft CAPE (DCAPE) in their

case was computed as 900 𝐽 𝑘𝑔−1 despite the presence of the near-surface stable layer. They

also note that the storm produced temperature rises of 5-8 °F, which correlated with the

DCAPE parcel trace. The studies discussed here show that both tornadoes and severe surface

winds can occur from elevated convection, but we still do not have a full understanding of

how these processes occur.

3. Numerical Simulations and Elevated Convection

3.1 Weather Research and Forecasting Model Background

3.1.1 WRF History

The WRF is a NWP model designed for both research and operational applications. It

was a collaboration between various U.S. agencies and universities with the goal to develop a

next-generation NWP model for the purpose of mesoscale weather prediction. The WRF is

currently at Version 3.7 (as of November, 2015), with the original Version 3 having been

released in June 2008. The two dynamical solvers (cores) available for the WRF are the Non-

hydrostatic Mesoscale Model (NMM) and the Advanced Research WRF (ARW). The WRF-

NMM is run operationally as the North American Mesoscale Model (NAM), while the WRF-

ARW is primarily used in research settings (although it can be run in real time).

3.1.2 WRF System Overview

The WRF is an entire model system, compromised of data assimilation, preprocessing,

and dynamics solving components. The WRF Data Assimilation (WRFDA) is capable of

processing a wide array of observations, including radar reflectivity/velocity and GPS

precipitable water/refractivity data, among many others. The WRFDA has implemented

4DVAR, but has options to use other techniques (3DVar, Ensemble DA, Hybrid

3DVAR/Ensemble). Users can also avoid the WRFDA altogether by using an existing

(archived) dataset for initial and/or boundary conditions. The WRF Pre-processing System

(WPS) uses the initial conditions supplied by a dataset to setup the domain of the WRF using

three separate programs: geogrid, ungrib, and metgrid. Geogrid creates the domain using the

user specified domain size, horizontal grid spacing, vertical resolution, and topography data.

The result is a gridded “empty” domain file to which meteorological data can be interpolated.

Ungrib is a simple program that converts data from grib format to a format that is useable by

the WPS for interpolation to the domain. Lastly, metgrid takes the data decoded by ungrib

and interpolates it to the domain created by geogrid. The result is a series of metgrid files that

are used by the WRF dynamical solver as initial and boundary conditions. Using these

metgrid files, the user runs the program called “real” to generate the input files for the WRF.

With these input files created, the user initiates a WRF run, which will in turn generate raw

model output in netCDF format. The WRF has been optimized to run in parallel on multiple

nodes or processors. The ability to run the WRF on multiple processors simultaneously is

critical in producing output in a timely manner, particularly for real-time simulations. Once

output data is generated, the user must utilize other software to visualize the model output.

3.1.3 WRF Technical Details

The WRF is a very flexible model system with many available features. In this section

the WRF-ARW will be primarily discussed, bearing in mind that some details of the WRF-

NMM may be different. The WRF is a fully compressible, Euler non-hydrostatic model

(hydrostatic option is available). It uses the Arakawa C-grid staggering and a terrain-

following vertical coordinate system (called η; shown in Equation 2) where the vertical

coordinate is equal to:

(2) 𝜂 =(𝑝ℎ−𝑝ℎ𝑡)

(𝑝ℎ𝑠−𝑝ℎ𝑡)

Where 𝑝ℎ is the hydrostatic component of the pressure and the other two terms are the

pressure at the surface (𝑝ℎ𝑠) and at the top boundary (𝑝ℎ𝑡). This is a coordinate system

following Laprise (1992). Time integration is done using a 2nd

or 3rd

-order Runge-Kutta

scheme. A useful feature of the WRF is the ability to implement an adaptive time step. This

type of time step changes depending on the wind fields in the model. Throughout a

simulation, the WRF can compute the Courant number and change the time step such that the

model moves toward a target Courant number. This ability to maximize the time step during

a simulation is very useful in reducing the time and computational cost of a simulation.

A wide range of model physics are available in the WRF. There are 14 microphysics, 13

cumulus, 11 planetary boundary layer, 7 surface layer, 13 land surface, 5 longwave radiation,

and 6 shortwave radiation parameterization schemes. For explicitly resolved convection, no

cumulus parameterization is used (it is effectively “turned off”). The WRF can be run either

as a regional or global model with various nesting options available. Full details of the WRF

Version 3 are available in Skamarock et al. (2008).

3.2 Numerical Simulations of Convection

Following the research of Weisman et al. (1997), explicitly resolved simulations of

convection are advised to have horizontal grid spacing of 4 km or less. Conversely,

simulations with horizontal grid spacing in excess of 10 km are advised to use a cumulus

parameterization scheme. Simulations with horizontal grid spacing of kmxkm 104 are

not recommended, given the ambiguity of whether or not to use cumulus parameterization.

Also of interest is the so-called “effective resolution” of the simulation, or what scale of

phenomena it can simulate. Skamarock (2004) notes that the effective resolution of the WRF

model is approximately 7Δx, which gives a minimum resolution 28 km for explicitly

resolved convection since the maximum horizontal grid spacing should be 4 km. This finding

also applies to simulations with Δx greater than 10 km, thereby making the effective

resolution of a model with cumulus parameterization no better than 70 km. Simulations

without convection parameterization solve the full vertical motion equation (Equation 3):

(3) 𝑑𝑤

𝑑𝑡= −

1

𝑝

𝜕𝑝

𝜕𝑧+ 𝑓𝑢 cot 𝜑 − 𝑔 + 𝐹𝑧

where the LHS is the change in the vertical component of the wind with time and the RHS

has four terms which are: the vertical pressure gradient, the coriolis force, acceleration due to

gravity, and friction. This is distinct from hydrostatic models that neglect the coriolis and

friction terms because scale analysis reveals that they are small relative to the pressure

gradient and gravity terms. This simplification is called the hydrostatic approximation

(Equation 4):

(4) 0 = −1

𝑝

𝜕𝑝

𝜕𝑧− 𝑔 ⇒

𝜕𝑝

𝜕𝑧= −𝑝𝑔

This approximation assumes that vertical motion is zero, meaning small vertical motions are

diagnosed using the continuity equation. Weisman et al. (2008) compared the results of a 4

km explicit convection forecast with a 12 km forecast using convection parameterization.

The results showed that value was added by using the explicit convection simulations. Figure

3 shows examples of two model runs from Weisman et al. (2008), one using convection

parameterization (panel c) and one that does not (panel d). Figure 3 (c) shows precipitation

generated by the model due to instability being released by the convection parameterization

scheme. It is noticeably coarser than the output in Figure 3 (d) due to the larger grid spacing

used with convection parameterization.

Figure 3: The 0000-0600 UTC accumulated precipitation (mm) for 10 Jun 2005 from the (c)

Eta and (d) WRF-ARW 24-30-h forecasts. From Weisman et al. (2008).

3.3 Studies of Elevated Convection Involving Numerical Simulations

Some recent studies of elevated convection have utilized high-resolution numerical

weather prediction models to simulate past cases (e.g. Bryan and Weisman, 2006; Colby and

Walker, 2007; Tardy, 2007; Trier et al., 2006). Both daytime and nighttime cases have been

simulated, with the former and latter involving convection above a frontal surface (e.g. Colby

and Walker, 2007). The simulation of a storm producing severe surface winds by Bryan and

Weisman (2006) presented in section 2.3 used the model of Bryan and Fritsch (2002).

However, like most other elevated convection research, their study did not focus on the

fundamental driving processes, but rather on a result of elevated convection (in this case

severe surface winds).

4. Discussion and Future Work

There remain many unanswered questions surrounding elevated convection. Corfidi et al.

(2008) asked some of these questions:

1) Why is castellanus frequently banded, what determines the spacing of the

bands, and what factors influence the diameter of individual convective towers

within them?

2) Why do some elevated thunderstorms produce severe surface winds whereas

most do not?

3) What conditions govern the depth, strength, and longevity of elevated

convective clouds, and can these variables be observed and forecast?

4) Why do elevated supercells sometimes assume a linear arrangement?

5) Are elevated storms affected by storm outflow and surface cold pools? If so,

how?

6) Do elevated storms acquire rotation in the same manner as do surface-based

supercells?

7) How can supercells on the cool side of baroclinic zones produce tornadoes?

8) Is there a maximum limit to the depth of the cold air mass for elevated storms

to produce tornadoes?

9) How do supercells with most unstable parcels that do not originate at the

surface differ from supercells that are more purely surface based or are more

purely elevated?

Of these questions, some are not unique to elevated convection. For example, question 7

asks how supercells on the cool side of baroclinic zones can produce tornadoes. This question

presumes that we understand how tornadoes form in general, but don’t understand how they can

form in elevated environments. Before we can answer this question, our general understanding of

tornado development will need to be significantly enhanced. Question 9 raises a fundamental

question about elevated convection – simply put, are the processes within elevated storms

different from surface-based storms? Question 3 is also fundamental in the sense that we do not

yet understand why elevated convection arises in some cases – especially when it occurs away

from surface fronts. The questions raised by Corfidi et al. (2008) in combination with the

objectives of PECAN provide a useful starting point to begin further studies into the topic of

elevated convection.

5. Conclusion

PECAN has provided a unique and exciting opportunity to address the many unanswered

questions about elevated convection. Corfidi et al. (2008) has formalized many of these

questions in the literature, but there are additional questions that will also need to be addressed.

Chief among the questions will be investigating the mechanisms that lead to elevated convection.

The wide array of data collected during PECAN should provide the necessary information to

begin to address some of these questions

6. References

Augustine, J. A., and F. Caracena, 1994: Lower-Tropospheric Precursors to Nocturnal MCS

Development over the Central United States. Wea. Forecasting, 9, 116–135.

Bonner, W. D., 1968: Climatology of the Low Level Jet. Mon.Wea. Rev., 96(12), 833-850.

Bryan, G. H., and J. M. Fritsch, 2002: A Benchmark Simulation for Moist Nonhydrostatic

Numerical Models. Mon. Wea. Rev. 130. 2917-2928.

Bryan, G. H., and M. L.Weisman, M. L., 2006: Mechanisms for the production of severe

surface winds in a simulation of an elevated convective system. Preprints, 23rd

Conf. on

Severe Local Storms, St. Louis, MO, Amer. Meteor. Soc., 7.5. [Available online at

http://ams.confex.com/ams/23SLS/techprogram/paper_115224. htm.]

Colby, F. P., and B. E. Walker, 2007: Tornadoes From Elevated Convection. Preprints, 22nd

Conf. on Weather Analysis and Forecasting and 18th Conference on Numerical

Weather Prediction, Park City, UT, Amer. Meteor. Soc., 7A.8. [Available online at

http://ams.confex.com/ams/pdfpapers/124653.pdf.]

Colman, B. R., 1990a: Thunderstorms above Frontal Surfaces in Environments without

Positive CAPE. Part I: A Climatology. Mon. Wea. Rev, 118, 1103–1121.

Colman, B. R., 1990b: Thunderstorms above Frontal Surfaces in Environments without

Positive CAPE. Part II: Organization and Instability Mechanisms. Mon. Wea. Rev, 118,

1123–1144.

Corfidi, S. F., S. J. Corfidi, and D. M. Schultz, 2008: Elevated Convection and Castellanus:

Ambiguities, Significance, and Questions. Wea. Forecasting, 23, 1280–1303,

doi:10.1175/2008WAF2222118.1.

Geerts, B., D. B. Parsons, C. Ziegler, D. Turner, and R. Ferrare, 2013: Plains Elevated

Convection at Night (PECAN) Experimental Design Overview. 1–37.

Goss, S. M., R. L. Thompson, and E. Bookbinder, 2006: An elevated supercell with

damaging wind from the morning of 12 March 2006. Preprints, 23rd Conf. on Severe

Local Storms, St. Louis, MO, Amer. Meteor. Soc., 18.4. [Available online at

http://ams.confex.com/ams/pdfpapers/115238.pdf.]

Horgan, K. L., D. M. Schultz, J. E. Hales Jr, S. F. Corfidi, and R. H. Johns, 2007: A five-

year climatology of elevated severe convective storms in the United States east of the

Rocky Mountains. Wea. Forecasting, 22, 1031–1044.

Kincer, J. B., 1916: Daytime and Nighttime Precipitation and Their Economic Significance.

Mon. Wea. Rev, 44, 628–633.

Maddox, R. A., 1980: Mesoscale Convective Complexes. Bull. Amer. Meteor. Soc, 61,

1374–1387.

Maddox, R. A., 1983: Large-Scale Meteorological Conditions Associated with Midlatitude,

Mesoscale Convective Complexes. Mon. Wea. Rev, 111, 1475–1493.

Marsham, J. H., S. B. Trier, T. M. Weckwerth, and J. W. Wilson, 2011: Observations of

Elevated Convection Initiation Leading to a Surface-Based Squall Line during 13 June

IHOP_2002. Mon. Wea. Rev, 139, 247–271, doi:10.1175/2010MWR3422.1.

Moore, J. T., F. H. Glass, C. E. Graves, S. M. Rochette, and M. J. Singer, 2003: The

Environment of Warm-Season Elevated Thunderstorms Associated with Heavy Rainfall

over the Central United States. Wea. Forecasting, 18, 861–878.

Lackmann, G., 2011: Midlatitude Synoptic Meteorology. 1st edn. AMS.

Laprise, R., 1992: The Euler Equations of motion with hydrostatic pressure as an

independent variable. Mon. Wea. Rev., 120, 197–207.

Rochette, S. M., and J. T. Moore, 1996: Initiation of an Elevated Mesoscale Convective

System Associated with Heavy Rainfall. Wea. Forecasting, 11, 443–457.

Skamarock, W. C., 2004: Evaluating Mesoscale NWP Models Using Kinetic Energy

Spectra. Mon. Wea. Rev. 132, 3019-3032.

Skamarock W. C., J. B. Klemp, J. Dudhia, D. O. Gill, D. M. Barker, M. G. Duda, X. Huang,

W. Wang, J. G. Powers, 2008: A Description of the Advanced Research WRF Version

3. NCAR Tech. Note NCAR/TN-475+STR, 113 pp.

Tardy, A, 2007: Climatology and Forecasting Application for Elevated Thunderstorms in the

Great Basin and West Coast of the United States. Preprints, 22nd Conf. on Weather

Analysis and Forecasting. Park City, UT, Amer. Meteor. Soc. P2.12. [Available online

at http://ams.confex.com/ams/pdfpapers/ 124729.pdf.]

Trier, S. B., and D. B. Parsons, 1993: Evolution of Environmental Conditions Preceding the

Development of a Nocturnal Mesoscale Convective Complex. Mon. Wea. Rev. 121,

1078–1098.

Trier, S. B., C. A. Davis, D. A. Ahijevych, M. L. Weisman, and G. H. Bryan, 2006:

Mechanisms Supporting Long-Lived Episodes of Propagating Nocturnal Convection

within a 7-day WRF Model Simulation. J. Atmos. Sci, 63, 2437–2461.

Weisman, M. L., Skamarock, W. C., Klemp, J. B., 1997: The Resolution Dependence of

Explicitly Modeled Convective Systems. Mon. Wea. Rev., 125, 527-548.

Weisman, M. L., C. Davis, W. Wang, K. W. Manning, and J. B. Klemp, 2008: Experiences

with 0–36-h Explicit Convective Forecasts with the WRF-ARW Model. Wea.

Forecasting, 23, 407–437, doi:10.1175/2007WAF2007005.1.