Publication Abstracts
Abstract of the Doctoral Dissertation
Author: Kong, F.-Y.
A three-dimensional hailstorm numerical model with fully elastic primi-
tive equations is developed. The model contains more detail bulk-water parame-
terized microphysics, including 4 warm cloud microphysical processes, and 17
ice phase microphysical processes, in which the Hallett-Mossop process and the
secondary ice crystal production process owing to the shattering of relatively
large supercooled drops during freezing are also parameterized explicitly, and
the overall enhancement effect to the ice crystal number by all ice multiplica-
tion processes is considered by adaption of the concentration formula, In add-
tion, a surface layer parameterization scheme is joined into the model, so that
it is able to simulate the evolution processes of hailstorms over lower boundar-
ies with inhomogeneous thermal or roughness properties.
The model has been runned experimentally to evaluate its stability and
validity. some important features of 3D convective systems, such as weaker
compensating downdraft outside clouds, subcloud cold pool and cold outflow, weak
echo structure, vortex pair and its splitting, and middle-level barrier-flow
pattern, are successfully simulated.
Four aspects of microphysics and dynamics of hailstorms, and their
interactions have been investigated by using the numerical model:
(1) The impact features and relative importances of the 17 ice phase
processes parameterized in the model are numerically evaluated, the special
emphasis is placed on contributions of the ice multiplication processes. The
simulations show that: The influences of ice multiplication are largely depen-
dent upon environments and conditions of clouds themselves. For cumulus clouds
in relatively warm atmosphere, the production of secondary ice crystal promotes
ice glaciation in clouds, with the results that more latent heat released warms
the upper portion of clouds and causes higher cloud tops, the ground precipita-
tion is advanced and the amount of solid precipitation is largely increased.
Whereas for those clouds in cold environment, the ice multiplication processes
no longer make significant contribution to cloud glaciation, but can modify the
fractions of solid and liquid precipitation. For some long-lived convection
systems, the processes even affect the cloud's dynamics and macro-feature.
However, for isolated thunderstorms with shorter life cycle, their evolution
features can not be influenced obviously.
In the two processes considered, the shattering mechanism plays a
dominant role in the effects of the ice multiplication on the microphysical
structure of convections. The enhancement correction to ice crystal concentra-
tion generally intensifies the overall effects of multiplication processes.
(2) The numerical experiments are made to evaluate the sensitivities of
convective storms to the ice phase microphysical processes. It is found that:
With ice phase processes, the modeling storms develop more intensively, the
total released latent heat and the surface precipitation rates reach their peak
values earlier. Especially, when the storm environment is relatively colder and
the supercooled layer is very deep, the effects of ice phase processes can be
strong enough to modify the storm's dynamic structure and life cycle. Analyses
show that the modification is due to the differences of microphysics included
in the model affect the optimum 'buoyance-shear' equilibrium of the storm.
These results implicate that the neglect of ice phase microphysics in
numerical cloud models will underestimate the intensity and precipitation of
simulating storms, or even fail, in some colder environments, in simulating
correct storm structure and lifetime. If the processes are excluded in mesoscale
models, prediction errors will be generated by means of underestimating latent
heat contributions. The error in latent heat can range from several ten to
hundred percent in magnitude.
(3) The three- and two-dimensional numerical experiments are made to
incestigate the evolution and surface precipitation features of convective
storms in mono-dirrectional low-level wind shear environment. The simulations
show that: In order to trigger a convection in low-level wind shear environment
by thermal bubble, a more intensive or extensive temperature disturbance is
needed, whereas it can be initiated much easier by cold outflow. Furthermore,
low-level shears with certain intensity lead the peak intensity of convections
to decrease, but a longer lifetime, larger amount of cumulative surface precipi-
tation, and more extend rainfall area with some smaller peak rainfall rate. The
analyses of flux distribution of water vapor through cloud base level find that
low-level shear in certain intensity is favourable to maintaining a steady,
long-lasting supply of water vapor, so as to promote the lifetime and precipita-
tion of storms.
Although the uses of two-dimensional model in simulating convective
activities in mono-directional shear environment have some severe distortions,
the time evolution features of maximum updraft velocities and total surface
precipitation are coincident with the results of three-dimensional simulations
qualitatively.
(4) The influences of large area of semi-unbounded cold water surface
on the evolution, propagation, and precipitation or hail shooting of hailstorms
in summer afternoon have been simulated, using real sounding profiles for
temperature, humidity, and wind. The model has successfully simulated the
significant modification of the propagation path of hailstorms near cold water
surface. The way of path change can be either 'along-bank' or 'toward-bank',
according to positions of the systems relative to convergence zone of water-land
circulation. The simulations also show that: The hailstorms developing or propa-
gating within convergence zone of local circulation will be intensified, and
produce much heavier hail shooting, whereas those over cold water surface or
cold modification layer will be strongly inhibited.
On the orther hand, because the direction and intensity of low-level
wind govern the degree of cold modification air and convergence zone of local
circulation moving into inner land, it is significantly meaningful to the degree
and property of the influences of cold water surface on hailstorm activities.
When strong low-level wind blows from water surface to land, convective systems
over a fairly wide range of land along water bank will be inhibited by thermal
effects of cold modification air in low level, without presenting obvious path
change.
Return to Education Backgound
Author(s): Kong, F., M. Huang and H. Xu
A fully elastic three-dimensional cold cloud numerical model with 15
parameterized ice-phase microphysical processes, including ice nucleation,
sublimation, riming, dry- and wet-growth of hailstone and melting, is developed
in order to investigate the interactions of ice-phase microphysics with cloud
dynamics and microstructure. The parameterization scheme for the processes and
the numerical technique involving ice-phase calculation are described. The
model is experimentally run to evaluate its stability and validity. Some
important features of 3-D convective systems, such as weaker compensating
downdraft outside clouds, weak echo strucure, votex pair and its splitting,
and middle-level barrier-flow, are successfully simulated.
Return to Publication List
Author(s): Kong, F., M. Huang and H. Xu
Two ice crystal multiplication processes are parameterized in the
compressible three-dimensional numerical cloud model described in the first
part of this study. The formation of secondary ice crystals by shattering of
relatively large supercooled drops during freezing and by the Hallett-Mossop
mechanism is explicitly considered, and the ice crystal consentration is
modified. For isolated convective clouds with warm base, the simulations show
that the inclusion of ice multiplication processes causes insignificant
modification on the cloud dynamics and macro-features, the simulated cloud
only produces slightly more total ground precipitation whereas the maximum
precipitation rate is increased and the appearing time of ice crystals is
advanced. However, the processes do affect the microphysics of the cloud:
They promote ice glaciation in clouds, with the result that more latent heat
released warms the upper portion of the cloud and causes a higher cloud top.
In the two multiplication processes, the shattering mechanism plays a dominant
role in the ice multiplication of clouds. The relative importance of each
ice phase process is also analyzed.j
Return to Publication List
Author(s): Kong, F., M. Huang and H. Xu
The numerical experiments on sensitivity of two types of convective
storms to ice phase microphysical processes have been made by using a fully
elastic three-dimensional cloud model with a more detail ice phase parameteri-
zation scheme. The simulations show that, in general, inclusion of ice phase
processes in the model leads to more intensive modeling storms--the total
latent heat released and the total precipitation increase obviously, the cloud
top rises, and the maximum ground precipitation rate is reached earlier.
These effects of ice phase processes can be much more significant in some
relatively cold environments, in which they even modify the dynamic structures
and life cycle of simulating storms. The mechanisms and favourable conditions
of ice phase influences on storms are also discussed.
Return to Publication List
Author(s): Kong, F.
(will input soon)
Return to Publication List
Author(s): Kong, F., M. Huang and H. Xu
The influences of large areas of semi-unbounded cold water surface on
the evolution, propagation and precipitation production of thunderstorms are
simulated by using a fully elastic three-dimensional numerical hailstorm model.
Real sounding profiles for temperature, humidity and wind are employed. The
model has successfully simulated the significant modification of the propagation
path of thunderstorms near the cold water area. The path change can be either
'along-bank' or 'toward-bank', depending on the position of the storm system
relative to the convergence zone of the water-land circulation. The simulations
also show that thunderstorms developing or propagating within the convergence
zone of local circulation will be intensified and produce much heavier hail,
whereas those over cold water surface or the air that has been cooled by the
water will be strongly inhibited.
The influence of the cold water surface on thunderstorm characters is
largely dependent upon the direction and intensity of the low-level winds.
Return to Publication List
Author(s): Kong, F. and and Y. Qin
A convective cloud transport model, without chemical processes, is
developed by joining a set of concentration conservative equations into a two-
dimensional, slab-symmetric and fuuly elastic numerical cloud model, and a
numerical experiment is completed to simulate the vertical transport of ground-
borne, inert gaseous pollutant by deep thunderstorm. The simulation shoes that
deep convective storm can very effectively transport high concentrated pollutant
gas from PBL upward to the upper troposphere in 30 to 40 minutes, where the
pollutant spreads laterally outward with strong anvil outflow, forming an
extensive high concentration area. Meanwhile, relatively low concentration
areas are formed in PBL both below and beside the cloud, mainly caused by
dynamic pumping effect and sub-cloud downdraft flow. About 80% of the pollutant
gas transported to the upper troposphere is from the layer below 1.5km AGL.
Return to Publication List
Author(s): Kong, F., M. Huang and H. Xu
A two-dimensional, slab-symmetric microburst numerical model with very
fine spetial resolution is developed and used to simulate the formation and
evolution of the dry and wet microburst lines. Some satisfactory results are obtained. The structure and evolution features of the simulated wet microburst
lines are coincident with the observations quite well. The dry microbursts,
driven by the evaporative cooling of the ice crystal precipitating element fall-
ing in a dry adiabatic layer, only produce a little ground rainfall and tempera-
ture drops. A weak stable layer several hundred meters thick can significantly
block and weaken the dry microburst.
Return to Publication List
Author(s): Kong, F.
A two-dimensional, non-reactive convective cloud transport model is
used to simulate in detail the vertical transport and wet scavenging of soluble
pollutant gases by a deep thunderstorm system. Simulations show that for gases
with not very high solubility, a deep and intense thunderstorm can still rapidly
and efficiently transport them from boundary layer (PBL) up to mid and upper
troposphere, resulting in a local significant increase of concentration in the
upper layer and a reduction in PBL. Dissolution effects decrease both the
incloud gas concentration and the upward net fluxes. The higher the solubility
is, the more remarkable the decrease is. However, for very low soluble gases
(H<100 M/atm), the influences are very slight. In addition, the effects of
irreversible dissolution and aqueous reactions in drops on the vertical
transport of gaseous pollutants are estimated in extreme.
Return to Publication List
Author(s): Kong, F. and and Y. Qin
The vertical transport features of gaseous pollutants, with a negative
exponent profile of concentration, by different types of convective cloud
systems are numerically investigated by using a two-dimensional, reactionless
convective cloud transport model. The results show that an isolated, weak storm
is able to pump pollutant gas out PBL and transport it to the mid-troposphere,
whereas a deep, intense thunderstorm can very efficiently transport air
pollutants up to the mid and upper troposphere and laterally spread with anvil,
forming an extensive concentration surge layer at altitude of ten-odd kilometers
altitude. Each type of convective transport results in concentration reduction
in PBL. In a wind shear environment the transport efficiency of deep thunder-
storm significantly increases and the pollutants enter into clouds on the
downshear side at low-level and spread downwind in anvil layer. On the other
hand, for a cumulus cloud with plenty of liquid water, the gas dissolution
effect is increased, and the irreversible aqueous reactions, in extreme, may
significantly weaken the vertical transports of pollutant gases with solubility
coefficents no more than 1000 M/atm.
Return to Publication List
Author(s): Kong, F. and J. Mao
A three-dimensional wind field analysis software based on the Beijing-
Gucheng dual-Doppler weather radar system has been built, and evaluated by using
the numerical cloud model producing storm flow and hydrometeor fields. The
effects of observation noise and the spatial distribution of wind field analysis
error are also investigated.
Return to Publication List
Author(s): Kong, F. and M. K. Yau
An explicit condensation scheme is developed for the Canadian Mesoscale
Compressible Community Model (MC2), which contains warm rain and ice-phase
microphysics processes. Three hydrometeor variables, cloud water mixing ratio,
rain water mixing ratio, and ice or snow mixing ratio, are explicitly predicted.
The effect of ice particle enhancement is parameterized in the scheme. Moreover,
the hydrometeor mass loading term is added into the model. A numerical simula-
tion test of the ERICA IOP2 winter cyclone case shows that the scheme is valid
with high efficiency. Despite running at the quite large time step comparable
to the MC2 system, it is able to generate reasonable magnitude and distribution
of the cloud and precipitation fields and to successfully capture the cyclone's
explosive deepening rate and general evolution feature without increasing too
much computing resource.
[Click HERE to download the entire paper (PS)]
Return to Publication List
Author(s): Kong, F. and M. K. Yau
A numerical study is conducted using the MC2 non-hydrostatic, semi-Lagrangian
semi-implicit limited-area model to investigate the evolution and structure
of the second strongest extratropical marine cyclone during the ERICA (IOP2).
Diagnosis of cyclogenesis based on the model output and some sensitivity tests
are also carried out to study the influences of deep convection on the storm's
explosive deepening process. For the later purpose, an explicit condensation
scheme containing warm rain and ice-phase microphysics processes is developed
within the MC2 framework. The control simulation successfully produces the
observed cyclone's evolution and structure characteristics. The sea level
central pressure falls 40 mb within 24 hours, 3 mb less than the observation.
The 30 hour simulated low center deviates from the observation by less than
75km. The cloud signature is excellently in agreement with the satellite
imagery. Besides the well simulated warm and cold frontal precipitation bands,
the coastal frontal precipitation accompaning IOP2 storm is also successfully
reproduced excepting some position deviation. Different condensation approaches
can generate quite similar deepening rate and storm track. However, the explicitmicrophysics scheme produces much stronger low level PV, especially along the
cold front zones, and leads to more fully seclusion of the cyclone center as
well. The inversion of PV anomaly of moist vs. dry run shows that the low and
mid level diabatic condensation contributes a major role to the cyclone's
deepening besides the horizontal thermal advection within the lower boundary.
[Click HERE to download the entire paper (PS)]
Return to Publication List
Author(s): Kong, F. and M. K. Yau
A high resolution simulation of the ERICA IOP2 explosive marine cyclone using
MC2 with a newly developed efficient explicit microphysics scheme has been
carried out. In addition to reproducing well the cyclone's explosive deepening
and its general characteristics, the simulation reveals many important fine
structures. The simulated bent-back front exhibits several mesoscale
perturbation, while both the primary and secondary cold fronts show very narrow
(about 20km) and sharp baroclinic features. Periodic mesoscale precipitation
cores are embedded along the cold fronts, with evidence of eddy-like and
hook-look structures. As cyclogenesis proceeds, the primary cold front and
the warm front gradually seperate from the bent-back front and the cyclone
center near the triple point. A second triple point forms between the
remaining bent-back front and the secondary cold front. The seperation
process is triggered at the low levels because of mesoscale downdrafts. The
deepest convections are found near the region of both triple points. Behind the
cold-front and in the area of the cold air outbreak, shallow cumulus cloud
streets, rotating cyclonically around the east side of the surface low
center, are successfully simulated. Moreover, A double spiral signature of the
cyclone center is evident in the low level vorticity field and hydrometeor
field. The explicit condensation scheme generates tremendous low-level PV
within the frontal zones, with a magnitude comparable to that calculated from
fine resolution observation data.
Return to Publication List
Author(s): Kong, F., V. Wong, and B. Zhou
Marine fogs and low ceiling stratus are the most dominant weather events
affecting aviation and navigation along the west coastal area of central
North America. An attempt has been made to simulate these phenomena using
a limited-area 3D mesoscale model -- the Advanced Regional Prediction System
(ARPS) developed at the University of Oklahoma. For the purpose several
modifications have been made to the model, includeing a new condensation
scheme which allows more realistic microphysics in low-supersaturation
environment like marine fogs and stratus, and a better treatment of MBL
processes. The simulations are initilized using 40km ETA model output
on the days with marine fog and/or low stratus off San Francisco Bay area
between May and June 1997. Various satellite imagery, especially the NRL
experimental GOES-9 nighttime and daytime fog and low cloud products,
along with surface observation data serve as verification dataset.
Preliminary results show that the ARPS model, with the modifications, is
able to simulate the formation of marine fogs under favourite large scale
circulation and sea surface thermal pattern. The persistent northerly
flow along the west coast and resulting large scale subsidence leads to
significant cooling and moistening within MBL, which is the major
mechanism in forming marine fogs. Among MBL physics processes, mixing helps
the MBL air reach saturation point, radiation cooling at fog top layer
increases inversion above MBL. With respect to fog distribution and
variation features, the model generated results agree reasonablly well
with the verification dataset. Furthermore, simulations are also carried
out to test the model's capability in predicting fog clearing and/or its
evolution into low stratus.
[Click HERE to download the entire paper (PS)]
Return to Publication List
Author(s): Kong, F.
This paper presents an experimental simulation of a summer season marine
fog-stratus case along the west coast of California using the U.S. Navy
COAMPS(tm) model. The purpose is to show the potential usefulness of mesoscale
models in forecasting this type of marine boundary weather phenomenon. The role
of data assimilation and the impacts of solar radiation, microphysics, and
vertical resolution in improving the forecasts are examined. The model
capability in forecasting the burn-off process over the San Francisco Bay
area is also tested with very high horizontal resolution (2 km grid size)
using the model's one-way nesting technique. The model demonstrates promising
capacity in this case to replicate the temporal and spatial cloud coverage
over the San Francisco Bay and surrounding area, shown in satellite imagery,
despite a 2 hour lag to complete clearing over the Bay. This study also
suggests that a better microphysics parameterization and proper representation
of microphysics in the solar radiation scheme are both important forCOAMPS(tm)
to produce more realistic simulations and to improve the burn-off forecast.
[Click HERE to view the entire paper (pdf)]
Return to Publication List
Author(s): Fiedler, B. H. and F. Kong
A new E-l boundary layer scheme is tested within the U.S. Navy's COAMPS
model. The goal is to give COAMPS the capability to simulate mesoscale
cellular convection. The new scheme is aimed to be consistent with both
classic results for clear entrainment and recent calibrations, derived
from large-eddy simulations, for entrainment into smoke clouds and water
clouds. A parameter is included in the scheme that allows sub-grid
transpot to be reduced so that, when the model has 2km grid spacing or
less, more of the transport is forced to occur in resolved convection.
At 2km grid spacing, the scheme allows COAMPS to simulate the break up
of a stratocumulus cloud deck into mesoscale cellular convection.
[Click HERE to view the entire paper (pdf)]
Return to Publication List
Author(s): F. Kong, K. K. Droegemeier, and N. L. Levit
Using a non-hydrostatic numerical model with horizontal grid spacing of 24 km
and nested grids of 6 and 3 km spacing, we employ the scaled lagged average
forecasting (SLAF) technique, developed originally for global and synoptic-scale
prediction, to generate ensemble forecasts of a tornadic thunderstorm complex
that occurred in North Central Texas on 28-29 March 2000. This is the first
attempt, to our knowledge, in applying ensemble techniques to a cloud-resolving
model using radar and other observations assimilated within non-horizontally
uniform initial conditions and full model physics. Our principal goal is to
investigate the viability of ensemble forecasting in the context of explicitly
resolved deep convective storms, with particular emphasis on the potential value
added by fine grid spacing and probabilistic versus deterministic forecasts.
Further, we focus on the structure and growth of errors as well as the
application of suitable quantitative metrics to assess forecast skill for highly
intermittent phenomena at fine scale.
Because numerous strategies exist for linking multiple nested grids in an
ensemble framework with none obviously superior, we examine several,
particularly in light of how they impact the structure and growth of
perturbations. Not surprisingly, forecast results are sensitive to the strategy
chosen, and owing to the rapid growth of errors on the convective scale, the
traditional SLAF methodology of age-based scaling is replaced by scaling
predicated solely upon error magnitude. This modification improves forecast
spread and skill, though we believe errors grow more slowly than desirable.
For all three horizontal grid spacings utilized, ensembles show both qualitative
and quantitative improvement relative to their respective deterministic control
forecasts. Nonetheless, the evolution of convection at 24 and 6 km spacings is
vastly different from, and arguably inferior to, that at 3 km because at 24 km
spacing, the model cannot explicitly resolve deep convection while at 6 km, the
deep convection closure problem is ill-posed and clouds are neither implicitly
nor explicitly represented (even at 3 km spacing, updrafts and downdrafts only
are marginally resolved). Despite their greater spatial fidelity, the 3 km grid
spacing experiments are limited in that the ensemble mean reflectivity tends to
be much weaker in intensity, and much broader in aerial extent, than that of any
single 3 km spacing forecast owing to amplitude reduction and spatial smearing
that occur when averaging is applied to spatially intermittent phenomena. The
ensemble means of accumulated precipitation, on the other hand, preserve peak
intensity quite well.
Although a single case study obviously does not provide sufficient information
with which to draw general conclusions, the results presented here, as well as
those in Part II (which focuses solely on 3 km grid spacing experiments),
suggest that even a small ensemble of cloud-resolving forecasts may provide
greater skill, and greater practical value, than a single deterministic forecast
using either the same or coarser grid spacing.
Return to Publication List
Return Home Page