Excessive Rainfall Discussion
NWS Weather Prediction Center College Park MD
800 PM EDT Sat Sep 21 2019

Day 1
Valid 16Z Sat Sep 21 2019 - 12Z Sun Sep 22 2019

...THERE IS A MODERATE RISK OF EXCESSIVE RAINFALL IN THE CENTRAL
PLAINS OVERNIGHT...

...Kansas / Missouri and Adjacent States...
In the 01Z update the Moderate Risk area remains fairly intact,
just having been backed up a little bit toward the north and west.
The event may be sluggish to unfold, as it is more reliant on deep
layer ascent rather than strong instability, and separate
corridors of low level convergence and mid level gradients will be
competing between northern Kansas and eastern Kansas overnight. If
convection congeals, however, it is most likely to do that within
the Moderate Risk area, which perhaps in spite the growth of
numerous outflows overnight, should see an expansion and
consolidation of convection the latter half of the night and first
part of Sunday in response to increasingly favorable upper jet
dynamics.

Our expectations best match the 12Z WRF-ARW and NSSL WRF, as well
as some recent HRRR runs. Wind profiles within deep layer
southwesterly flow are favorable for training, and outflow
boundaries could cause storm initiation to become anchored at
times. Again, though, because of broad inflow and initiation in
multiple areas, the event may be slow to ramp up toward producing
significant flash flooding. Ultimately, though, expect 1 to 3 inch
amounts to become common, with potential for spotty 5 inch amounts
per the hi-res models and the strong moisture anomalies in place.
The moisture is sourced both from the sheared-out remnants of
Tropical Storm Imelda and the high level contribution from ongoing
tropical cyclones over the eastern Pacific. Resulting PW value are
near the max observed for late September in this region, and
supportive of efficient rainfall processes.

Of particular concern is the upper flow fanning out toward 12Z as
the eastern Pacific moisture plume becomes more consolidated over
eastern Kansas and western Missouri per GFS 400-200 mb layer
fields. If sufficient instability remains, we could see flooding
perhaps escalate late in the night, continuing into Sunday morning.


...Oklahoma / Texas / New Mexico...
Rich moisture had advected westward to meet the approaching deep
layer trough structure over the southern High Plains. Convection
this evening was occurring in a very moist environment with
sufficient instability to support heavy rain rates. As of 00Z,
flash flooding had not been too much of a concern, given a
predominantly cellular mode and widespread rather than focused
initiation. Winds are, however, favorable to allow for merging and
brief training. Model QPF signal is especially heavy over far
southwest Texas and also western Oklahoma. The Oklahoma signal may
be related to a local maximum of low level convergence forecast
there during the low level jet cycle tonight. SSEO neighborhood
probabilities of rainfall exceeding flash flood guidance, and HREF
probabilities of 2-inch rainfall overnight are just about equal
within a zone from southwest Texas to southern Kansas, so we
maintained a narrow Slight Risk area.

Burke


Day 2
Valid 12Z Sun Sep 22 2019 - 12Z Mon Sep 23 2019

...THERE IS A SLIGHT RISK OF EXCESSIVE RAINFALL EXTENDING FROM THE
MID MISSISSIPPI ACROSS A PORTION OF THE UPPER GREAT LAKES AND
WESTERN OHIO VALLEY...

21Z Update - Based on 12Z model trends, shifted the outlook areas
farther to the northwest across the mid Mississippi valley and
Great Lakes regions.  Also extended the areas farther to the
southwest across eastern Oklahoma into the Red River region.

Across the Southwest, shifted the Marginal Risk area farther
southeast into southeastern Arizona and far southwestern New
Mexico.  Southwesterly flow drawing deeper moisture into the
higher terrain of the region may support some locally heavy
amounts Sunday afternoon/evening, otherwise the 12Z models
continued to signal little potential for widespread heavy
amounts/flash flooding across the Southwest through Day 2.

Previous discussion -

...Mid Mississippi Valley into the Upper Great Lakes and western
Ohio Valley...
Deep moisture and marginal instability focused on a frontal
boundary slowly crossing the Upper Great Lakes, western Ohio
Valley and the Mid MS Valley is expected to support convection
capable of excessive rainfall during Day 2. There is still some
model spread concerning how quickly the front drops south, which
results in some north/south differences in the placement of the
highest rainfall amounts. In an attempt to mitigate some of the
differences in the location of the axis of highest rainfall, the
WPC QPF and Excessive Rainfall were based on a multi model blend,
including the 00z NAM CONUS Nest and the 00z WRF ARW. 
 
Convection is expected to be ongoing along the front extending KS
into northern MO during the first part of Day 2, as a surface wave
crosses the region. Behind the surface wave, the front is expected
to drop southeast ahead of a long wave trough crossing the
Northern and Central Plains. A 20/30 knot low level southwest flow
continues to supply 1.75/2.00 inch precipitable water air (which
is between two and three standard deviations above the mean)
across the region. Model soundings from the 00z GFS/NAM showed an
axis of 500/1000 J/KG of MUCAPE, as cloudiness from the Day 1
convection could tamp down the instability available for storms
during Day 2. The newly developing storms could become better
organized in the difluence associated with the right entrance
region of a 110 knot jet streak crossing the Upper Great Lakes. 
 
The depth of moisture in the column could support hourly rainfall
rates between 1.50/2.00 inches, especially early on across
portions of central MO into central IL, where training occurs
along the front. An approaching long wave trough should provide
enough mid level flow to keep the convection moving (after an
initial training before 22/18z), which could limit the threat for
additional training. The highest rainfall amounts were stretched
from central MO across central IL into far southwest MI, based on
a slowly progressive frontal boundary. 
 
Three hour flash flood guidance across this area is generally
between 1.50/2.50 inches, with the lowest values across northern
IL. Given the potential for at least some training along the front
in a high moisture content airmass, a Slight Risk was extended
from southeast KS/northern and central MO across central IL into
far southwest MI. Since there is some model spread with respect to
the placement of the front, a slower frontal movement could result
in training occurring in places that receive heavy rainfall during
Day 1. If this occurs, a Moderate Risk across portions of eastern
KS/northern MO could be needed in later forecasts.


...Arizona...
Short wave energy crossing the northern Baja Peninsula could
interact with moisture from Lorena to produce convection capable
of producing heavy to locally excessive rainfall over portions of
southern AZ during Day 2. However, the amount of instability
available to the lift and moisture is uncertain, which could have
an impact on the coverage of any storms and the extent of the
flash flood threat. Because of this, the WPC QPF and Excessive
Rainfall Outlook were based on a multi model blend. 
 
Subtle short wave energy crosses the northern Baja Peninsula in
the mid level flow ahead of a developing long wave trough is
expected to provide synoptic scale ascent across much of central
and southern AZ during Day 2, mostly after 23/00z. Ahead of the
short wave energy, a weak low to mid level southwest flow
transports 1.75/2.00 inch precipitable water air (which approaches
three standard deviations above the mean) from Lorena across
southwest AZ, peaking between 23/06z and 23/12z. The combination
of moisture and lift should support the development of widely
scattered convection across portions of southern AZ during this
time period. 
 
Model soundings from the 00z GFS/NAM showed 500 J/KG of MUCAPE
ahead of the short wave, which should be enough to allow deeper
convection to form. However, synoptic cloudiness associated with
the short wave could limit the amount of instability prior to the
arrival of the best moisture and lift, limiting the instability
available for deeper convection. This is reflected in the QPF
fields from much of the 00z guidance, suggesting that instability
could indeed the limiting factor for a flash flood threat. 
 
A Marginal Risk was placed over much of southern AZ for Day 2.
However, the model support for this has diminished over the past
24 hours (with the exception of the 00z UKMET), and if this trend
continues, the Marginal Risk could be dropped in later forecasts.

 
Pereira/Hayes

Day 3
Valid 12Z Mon Sep 23 2019 - 12Z Tue Sep 24 2019

...THERE IS A SLIGHT RISK OF EXCESSIVE RAINFALL OVER PORTIONS OF
CENTRAL AND SOUTHERN ARIZONA...

21Z Update -  There remains a multi-model signal for developing
heavy rainfall across portions of the Southwest as an amplifying
upper trough/closed low drops into the Southwest and begins to
interact with moisture associated with Lorena.  Unfortunately,
significant model spread and poor run-to-run continuity in the 12Z
models with respect to the timing/amplitude of the upper trough
did little to bolster forecast confidence.  However, did discount
the NAM with some certainty.  With a strong ridge upstream, its
more progressive solution was considered less likely.  Opted
instead for a QPF blend of the 12Z GFS and 00Z ECMWF - which
continued to show heavier amounts across central Arizona centered
along the Mogollon Rim - supporting the continuation of a Slight
Risk across the region.  However, the 12Z ECMWF has trended faster
with the upper trough and is now much wetter farther to the south
than its previous run.  At this point, it is on southern edge of
the guidance with its heavier amounts.  However, should other
models show similar trends, adjustments to the Slight Risk may be
forthcoming.  Farther east - supported by model trends and in
coordination with WFO Albuquerque, the Marginal Risk was removed
across much of New Mexico.  


Previous Discussion -

...Southwest...
Forcing associated with a closing mid level low over CA/Southwest
during Day 3 interacts with deepening moisture from Lorena to pose
a flash flood threat across portions of the Southwest states.
While there is some spread concerning how quickly the mid level
system closes off, there is a multi model signal for heavy
rainfall, especially across portions of southern and central AZ.
Based on this, the WPC QPF and Excessive Rainfall Outlook were
based on a multi model blend, with more emphasis placed on the 00z
NAM/UKMET/GFS.

Short wave energy crossing into the base of an evolving positively
tilted long wave trough along the West Coast begins to carve out a
closed mid level low, which drops southeast into western AZ by
24/12z. Ahead of the closing mid level low, difluence produces
increasing lift in the column, especially after 24/00z across
southeast CA into western AZ. Moisture from Lorena is drawn north
into the mid level circulation, with precipitable water values
reaching 1.50 inches (which is between two and three standard
deviations above the mean) being transported across much of
western and central AZ, peaking between 24/06z and 24/12z. The
combination of forcing (augmented in the left entrance region of a
90 knot jet streak crossing southern NV into southern CA) and
moisture should be sufficient to produce bands of heavy rainfall
across central and southern AZ, mainly after 24/00z.

The timing of the moisture and best lift (generally after 24/00z)
is expected to occur after the diurnal maximum of instability,
which could be affected by synoptic scale clouds associated with
the closed mid level low. In fact, model soundings show the best
instability near the AZ/southeast CA border before that time.
However, falling heights ahead of the mid level system should
result in steepening lapse rates, and 00z NAM model soundings
indicated the presence of 500/750 J/KG of MUCAPE (with the base of
the instability near 750 mb). This amount of instability, in the
presence of strong synoptic scale forcing, should be sufficient to
support embedded convection within the bands of heavy rainfall,
again peaking after 24/00z.

Given the amount of moisture in the column, hourly rainfall rates
between 0.50/1.00 inches are possible, especially where training
occurs in the mid level southwest flow (as propagation vectors
become better aligned with the flow). There is a multi model
signal for local 2.00/3.00 inches of QPF across south central AZ
(including members of the 00z University of AZ Regional WRF
models), which could pose a flash flood threat, particularly over
burn scars and dry washes. After collaborating with WFOs
TWC/PSR/FGZ, a Slight Risk area was placed over much of south
central and central AZ for Day 3.


...Northeast...
As a mid level closed low develops over upstate NY during Day 3,
moisture and marginal instability ahead of its associated cold
front could fuel convection that produces heavy rainfall across
portions of the Northeast. Ahead of the front, a low level west
southwest flow transports 1.50 inches (which is between two and
three standard deviations above the mean) in a fairly narrow band
across the Northeast, centered on 23/18z. There is some model
spread concerning how much instability will be available to
support convection, with 00z NAM model soundings across NY into
VT/NH and western ME showing the potential 500/1000 J/KG of MUCAPE
just ahead of the front. The faster 00z GFS shows very little in
the way of available instability, so instability could end up
being the limiting factor when it comes to the amount of
convection ahead of the front.

The depth of the moisture is a concern with respect to heavy
rainfall, but considering the uncertainty with respect to
convective coverage, it is difficult to gauge how organized any
flash flood threat might be. Much of the 00z guidance does show
the potential for local 1.00+ inch rainfall amounts, and given the
expected moisture, these numbers seem plausible. Three hour flash
flood guidance values are generally above 2.50 inches, so at this
point, it would seem the flash flood threat is very low. Based on
this, no excessive area was assigned for Day 3. However, if later
model solutions indicate a more cohesive convective threat, a
Marginal Risk could be needed in later forecasts.

Pereira/Hayes



Day 1 threat area: https://www.wpc.ncep.noaa.gov/qpf/94epoints.txt
Day 2 threat area: https://www.wpc.ncep.noaa.gov/qpf/98epoints.txt
Day 3 threat area: https://www.wpc.ncep.noaa.gov/qpf/99epoints.txt