A Comparison of the Thermodynamic and Kinematic Features in Recent Blizzard and Near-Blizzard Events in Iowa

The history of operational meteorology is shaded with many weather forecasting techniques with varying degrees of veracity and reliability. For centuries, the idea of a ring appearing around the sun or the moon was used to portend rainfall or snowfall. We know now that this phenomenon is one of the more reliable visual cues in forecasting and occurs with cirrostratus formations, often in advance of the warm front associated with extratropical cyclones. The maturation of meteorology as a science in the 20^th Century saw a departure from purely observational cues used in forecasting. However, the introduction of standardized observations and rapid communications permitted real-time weather analysis, which includes sea-level pressure analysis using isobars.

Of particular interest in this study is the forecasters' regional (i.e. Iowa) rule-of-thumb technique that four isobars (at a 4-hPa spacing) present within the political borders of Iowa during a wintertime, snow-bearing, extratropical cyclone is sufficient to create the wind speeds found in a blizzard (Schwartz and Schmidlin 2002). Such a pressure gradient would allow for winds leading to the significant blowing of falling snow, as well as the lifting and stirring of previously fallen snow, which under non-precipitation conditions is sometimes termed a “ground blizzard.” Indeed, a pattern of four isobars that are oriented north-to-south across Iowa (which is ∼500 km across, east-to-west), and assuming an atmospheric density of 1 kg m⁻³ at 42°N., results in a wind speed of 24.5 m s⁻¹ (∼48 knots) if friction is neglected.

As such, the purpose of this research is to determine the veracity of this rule and continue the examination by Moyer (2010) regarding the dynamic and kinematic features that attend blizzards and near-blizzards over Iowa. Key findings for the surface in Moyer (2010) blizzard cases included low pressure centers with pressures < 990 mb, 6-hour pressure falls of ≥ 5 mb, and pressure gradients ≥ 20 mb across Iowa. Father aloft, negatively tilted troughs appeared to dominate at 500 mb.

This paper will examine the “Four Isobar Rule” using composites of several recent blizzard cases. In section 2 we detail our data and methods. Section 3 provides composite results, and conclusions are offered in section 4.

In the process of conducting this study, it was necessary to identify blizzards in the state of Iowa. As such, we begin with a brief discussion of the National Weather Service criteria for a blizzard, which are as follows:

• Surface (10-meter) winds of 30 knots (15.4 m s⁻¹) or greater,

• Sufficient snow in the air to reduce horizontal visibility to less than ¼ statute mile (0.4 km), and

• The two conditions stated above expected to be met for 3 hours or more.

Thus, significant wind must be blowing snow around as it falls, or raising snow off the ground, and keeping that snow suspended and blowing it about for an extended period of time. Extratropical storm systems that met these criteria were placed into the Blizzard category. The comparison group of storm systems were termed Near-Blizzard storms, and were identified thusly:

• Either one of the above Blizzard conditions are met, or

• Both Blizzard Conditions are met, but for less than 3 hours

Dates of Blizzard and Near-Blizzard cases were acquired for the state of Iowa for the period January 1999 through January 2010 via the National Centers for Environmental Information's (NCEI) and its World Wide Web-based Storm Events Database (http://www.ncdc.noaa.gov/stormevents/ ). From these dates, surface maps of mean sea level pressure observations were generated and analyzed subjectively, with a 4-hPa isobar spacing. Subjective analysis is a manual means of data assessment, where contours of pressure are drawn by hand. In this way, we were able to test the Four Isobar Rule, the origins of which are not clear.

In order to complete more robust analyses on these Blizzard and Near-Blizzard cases, additional data were obtained from the North America Regional Reanalysis (NARR) database for the times when the Blizzard and Near-Blizzard events were at their peak (as indicated in the Storm Events Database) or the nearest synoptic time thereafter (0000, 0300, 0600, 0900, etc Universal Time Coordinated [UTC]); NARR data are only available every 3 hours UTC. The NARR datasets are gridded, three-dimensional renderings of the atmosphere, produced by continuous assimilation of actual, observed atmospheric data into a numerical weather prediction system (Mesinger et al. 2006).

The NARR datasets were then subjected to the storm-relative compositing approach of Moore et al. (2003). In each class of cases (Blizzard and Near-Blizzard alike), the composite grids were centered on the location where the Blizzard or Near-Blizzard report originated. Composites were then compiled using the compgem software, obtained from Saint Louis University. Although the Blizzard and Near-Blizzard observations came from all over Iowa, composites were displayed on a map centered on Des Moines, Iowa (41.53° N latitude and 93.67° W longitude) for the purposes of scale. Composite geopotential heights and sea level pressure, temperatures, winds, and derived fields (isotachs, divergence, vorticity, etc.) were examined at the levels of surface, 850 hPa, 500 hPa, and 250 hPa, to further support a more complete synoptic evaluation of Blizzard and Near-Blizzard cases (e.g., Market et al 2004; McCoy et al. 2017; Kastman et al. 2017). We also created Skew-T log p, divergence, and vertical motion profiles to further investigate thermodynamic and kinematic behaviors in the blizzard and near-blizzard environments. Subjective examination of each sounding suggested that the feature-preserving composite approach of Brown (1993) was not necessary here.

From the NCEI online archive, 15 Blizzard cases were identified along with 23 Near-Blizzard cases between January 1999 through January 2010. Data on mean sea level pressure were then obtained and subjectively analyzed (with a 4-hPa contour interval) for each of these events. By this analysis 13 of the 15 (87%) Blizzard cases met the criteria of four isobars crossing some part of the state of Iowa. The Near-Blizzard collection only had 6 out of 23 (26%) cases that met the “Four Isobar Rule” over Iowa.

Further analysis of these mean sea level pressure maps revealed a primary grouping of cyclone centers northeast (N=7) of the location of the 15 total Blizzard events (Fig. 1a). The remaining 8 events were scattered about the center of their parent extratropical cyclone. We will return shortly to the group of 7 cyclones for the purposes of rendering a composite analysis.

Of the 23 Near-Blizzard cases, there were two primary groupings: with a cyclone center to the southeast of the Near-Blizzard conditions (N=7; Fig. 1b), and those with a cyclone center to the southwest (N=8; Fig. 1c). As with the Blizzard cases, the remaining 8 Near-Blizzard events were scattered about the center of their parent extratropical cyclone. With the Blizzard and Near-Blizzard case groupings, the patterns are sufficiently frequent to invite further study, this time in the context of composite analyses.

We also note that, of the 15 Near Blizzard cases retained for further study, seven of them did feature Blizzard conditions in a neighboring state. This fact serves to underscore the placement of a storm with respect to Iowa. In other words, the Four Isobar Rule is more important to Iowa owing to its unique dimensions in determining the surface pressure gradient.

The typical Blizzard case composites shown here are associated with mature extratropical cyclones, as demonstrated in the upper troposphere by a negatively-tilted trough at 250 mb (Fig. 2a) along with a warm pool in the base of that trough. This is a signature of a tropopause undulation (e.g., Hirschberg and Fritsch 1991), further supported by the >90 knot (46 ms⁻¹) jet stream flow along the southern periphery of the warm pool. The flow at 500 mb (Fig. 2b) reveals a well-defined vorticity maximum in the base of the trough whose negative tilt is even more pronounced than the one at 250 mb. At 850 mb (Fig. 2c), northwest flow and significant cold advection are depicted over most of Iowa, with winds in excess of 40 knots (21 ms⁻¹) across the western third of the state. The composite sea-level pressure map (Fig. 2d) shows four isobars aligned north-south across Iowa, thus supporting the forecasters' rule of thumb. Indeed, the composite surface cyclone center is found northeast of Iowa, over the upper peninsula of Michigan. Below-freezing temperatures, cold advection, and west-northwest winds of ∼20 knots (10 ms⁻¹) blanket Iowa at the time of Blizzard initiation.

Near Blizzard cases were composited following the categories identified previously. Near Blizzard (SE) cases (Fig. 1b) were composited as for Blizzard cases (Fig. 2) and the results are shown in Figure 3. Unlike the Blizzard cases, the composite Near Blizzard (SE) cases are in an earlier state of their extratropical cyclone life cycle. The 250-mb analysis shows a more zonal flow over Iowa and a weaker, positively-tilted trough (Fig. 3a). The exit region of a jet streak is also in evidence at 250-mb, and we will see shortly that central Iowa, in particular, was home to divergence at this level at this time. A more narrow, elongated circulation is depicted at 500 mb (Fig. 3b) in these cases than in Blizzard cases. The flow is less amplified, and there is the suggestion of a positively-tilted trough. This positive-tilt signature is especially in evidence at 850 mb (Fig. 3c), associated with a cold frontal zone over central Iowa at this time. In keeping with previous analysis composites, the surface chart (Fig. 3d) reveals a developing extratropical cyclone, with the composite cyclone center just southeast of the location of the Near Blizzard conditions. Composite winds were weaker in these cases, and nearly half the state was covered in 2-m air temperatures > 0°C.

Near Blizzard (SW) composite cases (Fig. 4) show less upper air structure, with nearly zonal flow at 250 mb and limited evidence of a trough or significant jet features (Fig. 4a). An even smaller and weaker circulation is present at 500 mb (Fig. 4b), with a shallow trough to the west. This feature does provide weak positive vorticity advection over most of Iowa. A closed low appears in the 850-mb composites (Fig. 4c), but is broad and positioned vertically under the circulation at 500 mb (Fig. 4b). Yet, warm advection may be inferred over most of Iowa at 850 mb (Fig. 4c). Indeed, when looking at the flow at the surface (Fig. 4d), 850 mb (Fig. 4c), and 250 mb (Fig. 4a), significant veering in the wind profile evinces deep warm advection in the column over Iowa. However, the surface analysis shows a modest cyclone, again beneath the circulation centers above it with a weak pressure gradient and modest 10 knot (5 ms⁻¹) winds.

Vertical profiles from within the composite fields were also created for each type of event. For the Blizzard cases, the composite thermodynamic profile (Fig. 5) reveals the deep signatures of a lowered tropopause (∼400 mb) and subsidence inversion (∼700 mb). Given the synoptic setting depicted in Fig. 2, this sounding (Fig. 5a) is to be expected. What is unique here is the deep layer (∼100 mb) near the surface where the atmospheric temperature profile is nearly dry adiabatic, and capped by winds of nearly 40 knots (20 ms⁻¹). This composite column is also dominated by deep convergence (Fig. 5b) and lower tropospheric downward vertical velocities of +1 to +2 μb s⁻¹ (Fig. 5c) at both the level of the subsidence inversion but especially in the planetary boundary layer. Thus, synoptic scale convergence and descent occurred in the presence of a dry adiabatic lapse rate in the planetary boundary layer.

The composite thermodynamic profile for the Near Blizzard (SE) cases (Fig. 6) also mirrors well its composite setting (Fig. 3). The lower troposphere is more moist than the Blizzard cases, with a much more shallow planetary boundary layer though it has a dry adiabatic lapse rate and a weaker wind profile (Fig. 6a). Indeed, the backing wind profile in the lowest ∼150 mb supports the plan view suggestion of cold advection, northwest of the surface low. The composite profile for the Near Blizzard (SE) cases also demonstrates a more dynamic atmosphere, dominated by lower-tropospheric convergence and upper tropospheric divergence (Fig. 6b) and deep ascent ( -3 to -5 μb s⁻¹) up to 400 mb (Fig. 6c). Such profiles can produce significant snowfall, but these fail to support the blizzard criteria.

The vertical composite signatures for Near Blizzard (SW) cases (Fig. 7) are similar to those for their Near Blizzard (SE) counterparts. The column is again more moist than for Blizzard cases (Fig. 7a), although there is a deep layer of veering in the wind profile. This behavior supports the idea of warm advection at 850 mb discussed previously. Moreover, the thermal profile departs a bit more from the dry adiabatic lapse rate than the Blizzard or Near Blizzard (SE) cases. Low-level convergence and upper-level divergence are even more pronounced in these cases (Fig. 7b) leading to the classic “bow string” signature on the mid-tropospheric vertical motion profile (Fig. 7c). Again, these profiles are conducive to significant snowfall, but without a deeper dry adiabatic lapse rate in the planetary boundary layer and a stronger wind profile, blizzard criteria were not met in these cases.

It is clear from each of these composites that significant extratropical cyclones attend and support the blizzard and near-blizzard conditions therein. For example, similarities can be found between the blizzard composite shown here (Fig. 2a-d) and the 1987 blizzard, documented by Schneider (1990), which affected eastern Iowa. In that event, a deeper surface cyclone also matches well with a deep, negatively-tilted 500-mb trough. Another example wouls include the work of Martin (1998), who also studied a system (his Fig. 3), which would best qualify as a near-blizzard (NBSE; Fig. 3a-d) in this study.

The composites presented here support the assertion that when four isobars are present on the surface analysis across the state of Iowa in the wintertime, blizzard conditions are likely to be met. The upper air analyses depict deep troughs, strong winds, and a positive vorticity advection field, which suggest a well-developed cyclone. More importantly, these composite results are largely consistent (e.g., 850-hPa winds of 45 kt) with the findings of Moyer (2010). Near-blizzards are weaker, without the strength of a well-developed cyclone, and the isobars were not clustered sufficiently to induce the winds to attain the desired speed. There will be slight conflicts in the displays, due to the compositing process and a loss of detail in the averages. Overall, the forecasters' rule-of-thumb of having four isobars present is supported by this study.

The authors wish to thank Dr. Charles E. Graves and Dr. Chad Gravelle of Saint Louis University for the use of the SLUBREW meteorological compositing software.