On May 13, 2009, an outbreak of severe weather caused extensive damage across the Midwest, especially in Missouri. Three people were killed there by tornadoes, including two near Kirksville. In the image below, you can see the reports of tornadoes (red triangles) over the northern tier of the state. The post will review and analyze the factors that led up to this outbreak. First, we will examine the big picture or synoptic scale features that set the stage and primed the environment for an outbreak, then we'll look at the intermediate scale or mesoscale components that actually triggered and sustained the severe storms. Mesoscale features are generally between 2 km and 1000 km horizontally.
Figure 1: Storm Reports from May 13, 2009. Notice the trail of red triangles across Northern Missouri (circled in red) representing tornado reports. These tornadoes were associated with one long-lived supercell thunderstorm. The supercell and associated tornadoes were responsible for 3 fatalities. [Source: Storm Prediction Center]
THE BIG PICTURE
The air mass over Missouri changed significantly during the day and we'll examine the synoptic scale features responsible for those changes that ultimately set the table for the outbreak. We'll first review the environmental landscape at the surface at both 12Z and 21Z.
The surface map below shows several key features at 12Z. First is the southerly flow out of the Gulf over the Midwest. Surface temperatures and dewpoints are close to 70°F and 60°F respectively in the warm sector, below the warm front, which forms the boundary of this warm, moist air. At this time, our area of concern is north of the warm front.
Figure 2: Surface Map from 12Z May 13, 2009. Our area of concern (Northern Missouri) is north of the warm front at 12Z. [Source: Hydrometeorological Prediction Center (HPC)]
By 21Z, several things have changed; the approaching cold front is now into NW Missouri, and the warm front has lifted into Iowa. Northern Missouri should be experiencing surface warming and moistening, which can lessen Convective Inhibition (CIN) while enhancing the Convective Available Potential Energy (CAPE).
Figure 3: Surface Map from 21Z May 13, 2009. Notice the southerly flow out of the Gulf right up Red River Valley. Also note the warm sector between the approaching cold front and lifting warm front. [Source: Hydrometeorological Prediction Center (HPC)]
The 21Z analysis at 850-mb shows a strong low level jet stream conveying warm, moist air into Missouri. Also of note is the confluence of winds with 20-25 kt winds out of the NW over Nebraska joining with 30 kt winds out of the SSW over Missouri. These two factors at 850-mb are priming the atmosphere for deep convection. This low-level jet stream moistened and warmed the low levels (enhancing CAPE, eroding CIN), while also bringing strong low level winds out of the SSW, setting the stage for enhanced bulk shear.
Figure 4: 21Z 850-mb Analysis. A low-level jet stream conveys warm, moist air over Missouri. The dark shade of green indicates dewpoint temperatures greater than 14°C. Note that the area of concern is generally shaded dark green and between the 16°C and 14°C isotherms at 850mb (red dashed lines), suggesting saturation. [Source: Storm Prediction Center]
The dominate feature on the 500-mb chart is the curved jet streak over the northern plain states associated with a short wave trough. The key connection between a short wave trough with mesoscale features is upper level divergence east of the trough. Recall that divergence aloft supports convergence at the surface. This combination promotes upward motion and subsequently, cooling and steepening of the environmental lapse rate. So, as we look at this particular jet streak, we should expect divergence in the exit region.
Figure 5: 21Z 500 mb Analysis. The shaded regions indicate higher wind speeds at 500 mb, with the light aqua color representing winds over 80 kts and the Curved Jet Streak. Divergence aloft should be anticipated downstream of such a feature. [Source: Storm Prediction Center]
A quick look at the 12Z 300-mb analysis shows a shortwave trough over upper mountain west, with a broad area of divergence east of the trough. Also of note is the relatively light winds around 25-30 kts over Missouri.
Figure 6: 12Z 300 mb Analysis. The divergence downstream of the trough promotes upward motion from the surface and steepening of lapse rates. [Source: Storm Prediction Center]
By 21Z, though, the trough has deepened and moved eastward, increasing our upper level wind speeds and maintaining an area of upper level divergence.
Figure 7: 21Z 300-mb Analysis. Note the progression of the trough. Compare the wind speed and direction at 850-mb to those here. We should expect high bulk shear values. [Source: Storm Prediction Center]
Because the wind speed has significantly increased at a different vector than at 850-mb, we should also expect that the bulk shear values to be increasing. Bulk shear impacts the duration of storms by lessening the interference of downdrafts with rotating updrafts. The result is long-lived, rotating thunderstorms.
The conditions at 300-mb and 500-mb primed the area for deep convection, with upward motion and local cooling that eroded the CIN and enhanced the CAPE. Further, the location of the jet streak increased the velocity of winds at both levels, further enhancing the bulk shear.
CAPE & CIN Analysis
Thus far, we have seen synoptic features that enhance CAPE, while eroding CIN. The chart below from the SPC shows the work of the synoptic scale, with a tongue of high CAPE values over the Midwest and CIN eroding eastward. At 21Z, notice that the CAPE at Kirksville is near 2000 J/kg, while the CIN is close to just 25 J/kg. Even with weak CIN still in place, convection requires that parcels get a mechanical boost to the LFC. The approaching strong cold front should have enough lift to get parcels to the LFC. Once there, the heightened CAPE suggests strong updrafts are in order.
Figure 8: 21Z SBCAPE & SBCIN Analysis. Note the tongue of high CAPE values consistent with the low-level jet stream. Also note the erosion of CIN from west to east. The synoptic features are at work priming the atmosphere for deep convection. [Source: Storm Prediction Center]
Moisture Convergence Analysis
Surface moisture convergence increases upward motion and contributes to thunderstorm development. Recall our low-level jet stream in place at 850-mb and its southerly flow. That rich, warm, moist air is converging with a dry, cold air mass along the cold front described in the surface analysis. The result is rising, warm, moist air that contributes to deep convection. The chart below shows areas of moisture convergence in red contours. Notice an area of moderate convergence over Northern Missouri and Iowa. Moisture convergence affects storm mode, and in this case, with organized but only moderate convergence, we should expect thunderstorms to develop linearly, with some discrete cells possible.
Figure 9: 21Z Moisture Convergence Analysis. Moisture convergence is organized along the frontal boundary. Compare this chart with the 850-mb analysis and it's clear how the low-level jet stream is driving this moisture convergence. [Source: Storm Prediction Center]
The regional reflectivity composite below shows the storms as they tracked across the state. While some of the storms formed as discrete cells, note how they formed linearly, which confirms the effect of moisture convergence.
Figure 10: 2155Z Composite Reflectivity. Notice how the storm mode reflects the moisture convergence. [Source: Storm Prediction Center]
The radar reflectivity image below from the KEAX site shows a more detailed view of this line of storms. The supercell that produced the tornadoes is identified on the image. However, due to the distance of the cell from the radar, the radar cannot inspect the lower levels of the storm and much of the classic storm shape is hidden.
Figure 11: 2148Z KEAX Reflectivity. A closer view shows the line of discrete cells impacting the area. Note the distance between the radar site (bottom left) and the tornadic thunderstorm (top right). [Source: National Weather Service]
In order to detect rotation, Doppler radar measures wind speed and direction of particles. In the velocity image below, particles moving toward the radar are negative velocities (green) and particles moving away from the radar are positive velocities (red). The close proximity of strong positive and negative velocities indicates a velocity couplet. To detect tornadoes, we look for couplets with gate-to-gate shear of 90 kts or greater. In the image below, we can clearly see a velocity couplet, indicating a mesocyclone, or rotating thunderstorm. But because the storm is beyond 100km, we cannot confirm a tornado or a Tornado Vortex Signature (TVS). In this case, the rotation was associated with the Kirksville tornado, but the radar was picking up rotation well above the surface, which is insufficient evidence alone.
Figure 12: 2148Z KEAX Storm Relative Velocity. A switch of modes shows a mesocyclone in the Kirksville thunderstorm, but a TVS cannot be verified at these distances. [Source: National Weather Service]
The tornadic supercell that caused three fatalities in Northern Missouri on May 13, 2009 had its roots in the upper air pattern in place at the time. Because the shortwave troughs at 300-mb and 500-mb caused divergence over the region, upward motion at the surface fueled a low level jetstream at 850-mb and caused local cooling and steepening of the lapse rates. Further, the strong westerly winds associated with the jet streaks contributed to the bulk shear, which enabled long track storms. The low-level jet stream helped to moisten and warm the low levels, while causing deep moisture convergence along the cold front. Ultimately, the cold front and moisture convergence trigger the storms in the environment set up by the synoptic features.
Despite the work of the low-level jet stream and upper air divergence to erode the CIN, some CIN remained over the region, capping convection without lift. The cold front with significant moisture convergence lifted parcels to the LFC, where strong updrafts fueled severe thunderstorms. Without the bulk shear, a product of the intense upper level winds, these storms would not have been self sustaining. Thus, even after the storms initiated, the upper level environment played a role in sustaining the storms.