Petersen et al.

Flash floods associated with MCSs – observations

Petersen W. A., L. D. Carey, S. A. Rutledge, J. C. Knievel, N. J. Doesken, R. H. Johnson, T. B. McKee, T. Vonder Haar, and J, F. Weaver, 1999: Mesoscale and Radar Observations of the Fort Collins Flash Flood of 28 July 1997.

Bull. Amer. Met. Soc

.,

80

, 191-216.

Notable summertime floods of the western U.S.

Rapid City, S.D.

June 9, 1972 Big Thompson Canyon, CO July 31, 1976

The Fort Collins, CO Flood – July 30, 1997 5 fatalities, 40 injuries, half of CSU library collection ruined Most well documented flash flood – occurred close to CHILL research radar, two NEXRADS and over CSU DAS

Topography and radar locations

Christman field CSU

1600 27 July 1997- 1300 28 July 1997 1730 28 July 1997- 2300 28 July 1997

Coordinate system origin for later graphs (intersection of Taft Hill and Drake Roads)

Flood advisory issued Flooding north of FCL from earlier rains Flash flood watch issued Flood advisory cancelled NWS issues special statement outlining potential for flooding Urban Street Flooding Advisory issued Urban Operation Center Opened

5.3 inches/6 hr

Spring Creek Floods onto Roads Campus buildings ands streets flood Homes flooded Flash flood warning issued by NWS Train culvert breaks sending a wall of water into mobile home park Train derails at broken culvert

Ridge axis Dewpoint depression < 6°C shaded Temperature contours = 2°C Height contours = 30 m Weak flow aloft over Northern Colorado 500 mb Summer “monsoonal” flow – moist air aloft from SE Pacific

Cloud tops < -20°C each hour from 0600-1800 MDT Surface dewpoint > 60°F to east of line 500 mb vorticity centers Frontal positions at 0600 (a), 1200 (b), 1800 (c) and 2400 (d) MDT Note the easterly flow to the north of the surface front. This flow brought warm moist high q e air upslope to the foothills of the Rocky Mountains

1800 MDT sounding from Denver, CO overlaid on a sounding taken during TOGA COARE along the equator!

Atmosphere over eastern Colorado had characteristics of the tropical Pacific!

Extremely moist compared to climatology Not too unstable Other sounding parameters:

Radar echoes from flood region (red box) do not stand out as unusual Note this bow echo – this feature modified the strength of the easterly flow into the Fort Collins storm during part of its lifetime Flow and radar echoes at 18, 19, 20, 21 MDT

Low level easterly winds were sufficiently deep that low level clouds drifted from east to west

Wind barbs in knots (standard)

1416 MDT 1716 MDT Cloud drift winds

Cross section at 2 km AGL of radar echoes from Cheyenne, WY NEXRAD for period of most intense rainfall – Spring Creek is the black line Note strong echoes over Spring Creek basin and the redevelopment of convection and “training” of the heaviest rainfall on the south side of the complex.

Radial velocities from CHILL radar superimposed on reflectivity contours Radar beam Outflow from convection Inbound velocities Outbound velocities Cell regeneration region Spring creek

Cloud to ground lightning strokes between 1800 and 2300 MDT (different colors for each hour) (little box is flood area) Hardly any lightning!

Dual Doppler derived winds at 1 km AGL at 2110 and 2130 MDT Flood storm Bow echo to south It appears that the easterly flow into the flood storm accelerated as a result of some interaction with the storms within the bow echo, although the nature of the interaction was difficult to specify

Dual-Doppler derived u (easterly) wind component Easterly “jet” flowing toward convection over Ft. Collins Bookend vortex at northern end of bow echo

Strength of easterly wind (expressed as a deviation from a mean value for the time period) Rain mass flux determined with radar Amount of rain produced by the storm over Fort Collins highly correlated with the strength of the low level easterly inflow

Relationship of lightning strikes to mean reflectivity evolution in storm Rain mass flux Reflectivity -10°C Strong echoes in “warm rain” region, rather than associated with hail in ice region Lightning strikes Evidence from this data suggests that the warm rain process (collision coalescence) dominated precipitation production during much of the storm – making the storm more “tropical-like” and unlike the storms further east on the same day (see lightning chart)

Reflectivity profiles as a function of height, normalized to profile maxima for the Fort Collins storm (bold), tropical Atlantic convection (dashed) and the tropical west Pacific (thin).

Note similarity between tropical convection and the Fort Collins storm

ZDR – black KDP - Blue Vr – black Receding - solid LDR – blue Ice fraction - Black

Reflectivity (dBZ) Polarization radar variables show that precipitation process was warm rain process

Reflectivity (color) ZDR ( black) KDP (white) Microphysical processes suggested by polarization data: 2-3 mm raindrops formed by collision-coalescence rise through freezing level, freeze above about –10°C and fall to the northwest of updraft in heavy rain region.

Rainrate from Nexrad Z-R relation Rainrate from polarization algorithm Rainrates in flood conditions are estimated better with polarization radar compared to conventional Z-R relationships.

Blend of relationships NEXRAD standard Z-R

Values low and did not match spatial distribution

NEXRAD tropical Z-R

Values reasonable except lower value with CHILL matched spatial distribution

Multiparameter polarization estimate of precipitation

Slightly low maximum but best estimate across the gage network