1. Introduction
Mesoscale band formation within extratropical cyclones can dramatically affect the timing, intensity, and subsequent accumulation of precipitation. The effects of mesoscale bands are especially evident during the cold season, when snowbands can be responsible for snowfall rates exceeding 10 cm h−1 (4 in. h−1) and their limited scale can create extreme gradients in snow accumulation. Such characteristics have made mesoscale banding a challenge to diagnose and predict.
A variety of banding mechanisms have been proposed (e.g., Schultz and Schumacher 1999); however, previous studies have established that the primary mechanism for band formation on the poleward side of midlatitude cyclones is frontogenesis in the presence of weak moist symmetric stability (e.g., Thorpe and Emanuel 1985; Sanders and Bosart 1985; Nicosia and Grumm 1999). The assessment of frontogenesis and moist symmetric stability in an operational forecasting environment has been discussed by Wiesmueller and Zubrick (1998), Schultz and Schumacher (1999), and Nicosia and Grumm (1999). These studies have cited the need to identify the synoptic context in which favorable environments for band development are established.
Nicosia and Grumm (1999), in a study of three northeast U.S. snowstorms, showed that deformation north of the developing midlevel cyclones contributed to strong frontogenesis. The snowbands observed in their cases were coincident with this frontogenesis within a deep layer of weak moist symmetric stability, suggesting that the synoptic and mesoscale flow environment may influence mesoscale band formation (Fig. 1a). Novak et al. (2004) expanded on this work, providing a climatology and composite study of 48 single-banded events in the northeast United States. Single-band formation was found to be common in the comma-head portion of cyclones associated with cyclogenesis, as the development of a closed midlevel circulation supported deformation and associated frontogenesis northwest of the surface cyclone (Fig. 1b). These results are also consistent with the case study work of Martin (1998a, b), Banacos (2003), and Moore et al. (2005), who have documented similar synoptic and mesoscale flow evolutions in case studies of mesoscale banding in the central and eastern United States. These emerging conceptual models of the synoptic and mesoscale flow environments conducive to band formation are providing forecasters with an awareness of the potential for mesoscale banding 1–2 days in advance.
At the same time, new observational datasets and numerical prediction advances are improving forecast capabilities. The advent of high temporal and spatial resolution remote sensing datasets such as the Weather Surveillance Radar-1988 Doppler (WSR-88D) national radar network (Klazura and Imy 1993), satellites, wind profilers, and the Meteorological Data Collection and Reporting System (MDCARS; Moninger et al. 2003) have allowed forecasters to observe the formation and evolution of banded precipitation structures and their environments. Assimilation of these datasets into regional models such as the Rapid Update Cycle (RUC; Benjamin et al. 2004) and the National Centers for Environmental Prediction (NCEP) Eta Model1 (Black 1994; Rogers et al. 2001) is improving short-range (<12 h) forecasts. High-resolution (<12 km) limited-area models, such as the Workstation Eta, the fifth-generation Pennsylvania State University–National Center for Atmospheric Research (PSU–NCAR) Mesoscale Model (MM5; Dudhia 1993), and the Weather Research and Forecasting model (WRF; Janjić 2004), are providing explicit forecasts of mesoscale banded features. Finally, the advent of short-range ensemble forecast (SREF) systems (Stensrud et al. 1999) has allowed forecasters to gauge the uncertainty of forecast synoptic and mesoscale flow evolutions and to establish a level of confidence in model forecasts.
This paper draws on the above examples of emerging conceptual models of band development, new observational datasets, and numerical weather prediction advances to propose an ingredients-based, time- and scale-dependent forecast strategy to anticipate cold season mesoscale band formation within eastern U.S. cyclones. The focus of this paper is on anticipating the formation of precipitation bands that occur on the poleward side of eastern U.S. cyclones. For the purposes of this paper, a band is defined as the occurrence of a linear reflectivity structure > 250 km in length, 20–100 km in width, with an intensity > 30 dBZ maintained for at least 2 h; this definition corresponds to the single-band type in the band classification scheme proposed by Novak et al. (2004). The benefits and challenges associated with the proposed forecast strategy are illustrated through its operational application to the 25 December 2002 northeast U.S. snowstorm to show that accurate operational forecasts of mesoscale bands can be made based on our current conceptual understanding, observational tools, and modeling capabilities. The timing of this event is fortuitous in that it occurred only a few weeks after National Oceanic and Atmospheric Administration/National Weather Service (NOAA/NWS) forecasters received training on the proposed forecast strategy.
This paper is organized as follows. Section 2 contains descriptions of the datasets and products used to illustrate the forecast strategy. Section 3 outlines the forecast strategy, identifying ingredients for mesoscale band formation and delineating their assessment according to forecast projection. Section 4 illustrates the application of this strategy through an examination of the 25 December 2002 northeast U.S. snowstorm, in which intense mesoscale snowband formation was accurately predicted. Section 5 provides a discussion of the forecast strategy and a summary of the key points of this paper.
2. Datasets and products
A variety of observational and model datasets is utilized in the forecast strategy. Particular datasets used to illustrate the application of the forecast strategy to the 25 December 2002 northeast U.S. snowstorm include forecasts from the NCEP SREF system (Du and Tracton 2001), Global Forecast System (GFS), Eta and RUC models, and WSR-88D data. At the time of this case, the NCEP SREF system was run at a horizontal resolution of 48 km and was composed of 10 members, based on two different models and five different initial conditions for each model. The GFS was run at a horizontal resolution of spectral triangular 254 (T254; roughly equivalent to 55-km grid spacing) with 64 vertical levels within the first 84 h of the forecast (Environmental Modeling Center 2005). The Eta was run at 12-km horizontal resolution with 60 vertical levels (Rogers et al. 2001). The RUC was run at a horizontal resolution of 20 km with 50 vertical levels (Benjamin et al. 2004). WSR-88D mosaic radar data used in this case were composed of reflectivity returns from the 0.5° elevation scan from all radar sites, with 2-km spatial and 5-min temporal resolutions.
Archived 80-km horizontal resolution and 50-hPa vertical resolution display grids from the Eta and GFS, and 40-km horizontal resolution and 25-hPa vertical resolution display grids from the Eta and the RUC, were obtained from the Cooperative Program for Operational Meteorology, Education and Training (COMET). These display grids were available to NOAA/NWS forecasters in real time. To differentiate between model display resolutions, subscripts are used when referring to model fields (i.e., GFS80, Eta80, Eta40). The General Meteorology Package (GEMPAK; desJardins et al. 1991) version 5.6 was used to calculate and visualize kinematic parameters from the various model datasets. Frontogenesis is diagnosed using the simplified 2D form of the Miller (1948) frontogenesis equation [Novak et al. (2004); Eq. (1)], which isolates the frontogenetical effect of the horizontal wind.
NOAA/NWS forecast products are used in the forecast strategy to illustrate how the threat and occurrence of mesoscale band formation may be communicated to the public. A description of the NOAA/NWS products discussed in this paper appears in Table 1. Note that only product examples for the case of a mesoscale snowband are discussed; however, analogous forecast products may be used in the case of a mesoscale rainband.
3. Forecast strategy
a. Background
Ingredients-based forecast methodologies have been developed for warm season convective weather (McNulty 1978), heavy rainfall (Doswell et al. 1996), and winter precipitation (Wetzel and Martin 2001). These methodologies center on the three basic ingredients of lift, instability, and moisture. As previous research has shown, mesoscale band development is associated with frontogenesis in the presence of weak moist symmetric stability and sufficient moisture. In the proposed forecast strategy these three parameters—frontogenesis, weak moist symmetric stability, and moisture—are used as the ingredients for identifying mesoscale band formation. It is recognized that favorable microphysics [conditions leading to rapid snowflake growth for example; Roebber et al. (2003)] may enhance precipitation efficiency, and especially precipitation accumulation in the case of snowfall. However, microphysics is only considered to affect band intensity and is, thus, not considered an ingredient for mesoscale band formation in the proposed forecast strategy.
Since the predictability of an atmospheric feature is dependent on the scale of the feature [predictability decreasing as the scale of the feature decreases; Dalcher and Kalnay (1987); Droegemeier (1997)], the predictability of mesoscale bands will be limited to relatively short time scales. For instance, Zhang et al. (2003) have suggested that deterministic mesoscale precipitation forecasts are impractical beyond 2–3 days due to the upscale error growth associated with moist convective processes. In recognition of the predictability limits inherent to mesoscale bands, the proposed forecast strategy provides a time- and scale-dependent approach for assessing the band ingredients. Such a strategy is similar to the Snellman Funnel approach often used in the forecast process (Snellman 1982; Snellman and Thaler 1993). At extended forecast projections, the strategy focuses on assessing the possibility of cyclogenesis. At forecast projections of 24–48 h, the ingredients (frontogenesis, weak moist symmetric stability, and moisture) are assessed in the context of conceptual models proposed by Nicosia and Grumm (1999), Novak et al. (2004), and Moore et al. (2005) to determine whether banding is possible within the impending cyclone. As the event nears, the assessment of banding potential increasingly relies on the juxtaposition of the ingredients, providing a basis to specify with increasing detail when and where band formation will occur. This time and scale dependence are based on the premise that the synoptic-scale flow environment regulates the development of the subsynoptic-scale features (e.g., Schultz et al. 1998; Novak et al. 2004), which are inherently less predictable than the synoptic-scale flow in which they are embedded.
It is important to recognize that specific forecast projections contained in this strategy are intended as guidelines that can vary based on the predictability of the specific event. It is also important to appreciate that the forecast projections are subjectively based on the state of the art in conceptual understanding, observational datasets, and modeling capabilities at the time of this writing, and will likely change as advances are made in these areas.
b. Description of the forecast strategy
1) Greater than 48 h before the event
Climatological and case study results have demonstrated a dynamical link between the process of cyclogenesis, attendant deformation zones, and associated frontogenesis (Novak et al. 2004). The magnitude of cyclogenesis has an impact on the development and magnitude of deformation, the subsequent intensity of fronts, and the juxtaposition of dry and moist airstreams conducive to reducing the moist gravitational and symmetric stability (Nicosia and Grumm 1999). Current operational modeling systems, particularly ensemble forecast systems such as the NCEP Medium Range Ensemble Forecast system (Toth et al. 1997) and the NCEP SREF system, have demonstrated skill at indicating the possibility of a cyclogenesis event across a given region beyond 2 days. At these extended forecast projections the focus is not on the development of mesoscale bands, but rather if a cyclogenesis event (storm) is expected in the general region. Awareness of an impending cyclogenesis event can alert the forecaster that an environment supporting band formation may be present; however, the likelihood of band formation, let alone its timing and location, may have considerable uncertainty. As forecast confidence in the cyclogenesis event increases, a Winter Storm Outlook may be issued.
It is noted that although cyclogenesis often provides a favorable environment for band formation, bands may form in other synoptic situations given an environment characterized by frontogenesis in the presence of weak moist symmetric stability and sufficient moisture. For instance, Banacos (2003) discusses band environments associated with strong east–west-oriented frontal zones in the absence of significant cyclogenesis.
2) 24–48 h before the event
At forecast projections of 24–48 h, the band ingredients are assessed to determine whether banding is possible. Of particular interest is the development of midlevel deformation, which can be maximized north and west of the surface cyclone as the midlevel circulation develops, and in diffluent flow ahead of the surface cyclone (Fig. 1b). Deformation is important since it can contribute to frontogenesis when the angle between the isentropes and the dilatation axis is less than 45°, and frontogenesis serves as the mesoscale forcing for banded precipitation.
The relationship between the synoptic-scale flow evolution and the band ingredients affords forecasters an opportunity to assess if banding is possible 1–2 days in advance. Similar to predictions of other mesoscale phenomena such as lake-effect snow or severe thunderstorms, the identification of favorable synoptic environments can provide forecasters with a heightened awareness of the possible event occurrence. SREF systems can be used to gauge the uncertainty of forecast synoptic and mesoscale flow evolutions and to establish a level of confidence in model forecasts. For instance, spaghetti and probability charts (e.g., Sivillo et al. 1997; Toth et al. 2001) can be used to diagnose the uncertainty of synoptic and mesoscale features such as the strength and orientation of developing deformation zones, along with areas and amounts of heavy precipitation. Given sufficient confidence, a Winter Storm Watch may be issued for areas in which the potential for banding exists. However, at this time forecast projection uncertainty in cyclone and frontal positions limits explicit prediction of the timing and location of the banding event.
3) 12–24 h before the event
At forecast projections less than 24 h, confidence in the forecast synoptic-scale flow evolution is generally established. This level of confidence allows the forecaster to assess when and where band formation is favored by examining plan-view and cross-sectional environments of midlevel deformation zones and frontogenesis maxima in the model forecasts. Since frontogenesis, weak moist symmetric stability, and sufficient moisture have been identified as the primary band ingredients, the timing and location of the coincidence of these ingredients can outline the banded threat area. However, a number of challenges arise in the operational assessment of frontogenesis and moist symmetric stability.
Consider a conceptual model of the cross-sectional environment of a banded frontal zone (Fig. 2). Note that the frontal zone and the associated frontogenesis maximum slope toward the cold air. Since the frontogenesis maximum slopes, the level a forecaster chooses to assess frontogenesis in plan view can dramatically affect the area outlined for band development. In practice, assessment has focused on levels between 850 and 600 hPa (Banacos 2003; Schumacher 2003); however, the outlined location may vary by more than 200 km depending on which particular level or layer is chosen (Schumacher 2003). Observational studies have suggested that bands tend to form where a layer of weak moist symmetric stability lies just above the frontogenesis maximum (e.g., Trapp et al. 2001; Novak et al. 2004), as depicted in Fig. 2. These results highlight the importance of the coupling of frontogenetical forcing with weak moist symmetric stability in promoting band formation.
Once a threat area is outlined, the question remains whether the frontogenetical forcing is strong enough to support band formation. Operational experience has shown that no single threshold value can be correlated with band development, since moist gravitational and symmetric stability modulate the frontogenetical response, and other factors such as microphysics and moisture availability can alter precipitation development. An additional complicating factor is that the frontogenesis magnitude is dependent on the resolution of the computational grid (e.g., Schumacher 2003). Similar challenges arise when assessing the moist symmetric stability, as the calculation of saturation equivalent potential vorticity (EPV*) has been shown by Schultz and Schumacher (1999), Clark et al. (2002), and Jurewicz and Evans (2004) to be quite sensitive to model resolution and to the formulation of the EPV* in terms of the geostrophic wind (EPVg*) or the full wind (EPV*f).
Given these complicating factors, model vertical velocity and quantitative precipitation forecasts (QPFs) can aid in the assessment of band potential, providing evidence to support or refute the hypothesis of band formation. The hypothesis of band formation is supported when a narrow (relative to the resolution of the model) region of forecast ascent and precipitation corresponds to the forecast location of midlevel frontogenesis and weak moist symmetric stability (Schumacher 2003). Although the explicit timing, intensity, and placement of a model forecast precipitation band may be in question, the fact that the model dynamics are producing banded precipitation suggests the presence of a favorable environment. Precipitation fields from high-resolution models can provide additional insight into the nature (i.e., movement, intensity, and dissipation) of the precipitation event given the fine horizontal resolution and hourly accumulation periods of these fields. Similar benefits of using high-resolution model precipitation fields have been demonstrated by Roebber et al. (2002) for convective precipitation events. However, at 12–24-h forecast projections, the explicit timing, intensity, and location of the band should be considered as qualitative, rather than as quantitative, forecast guidance (Roebber et al. 2004).
The detailed assessment of the banding ingredients can aid in the decision whether to upgrade portions of Winter Storm Watches to Winter Storm Warnings. Wording within the warning text can specify where within the warning area the greatest threat exists for especially heavy snowfall, including mention of expected rates and total amounts. Graphical and gridded forecasts may be especially effective media to relate information regarding snowfall gradients and total amounts.
4) 0–12 h before the event
Within 12 h of the band event, short-range forecasts from models such as the RUC and Eta can be used in concert with observations to answer the critical question of when (within 3 h) and where (on the scale of counties; ∼100 km) band formation will occur. Similar to longer forecast projections, emphasis is placed on the forecast evolution of the midlevel frontogenesis, moist symmetric stability, and moisture; however, the temporal proximity to the event lends high confidence to the model forecast placement and magnitude of these ingredients. Additionally, short-range forecast model parameters are often available hourly, allowing for a detailed assessment of the forecast evolution of the event. Perhaps most importantly, the proximity to the event allows for comparison of observations with model forecasts, from which forecast updates can be made (Banacos 2003). For instance, developing areas of heavy precipitation can be compared with short-term model forecasts of frontogenesis, moist symmetric stability, and moisture fields valid at the analysis time. Based on this assessment, additional details can be added to Winter Storm Warnings and other forecast information. Short Term Forecasts may be especially useful to specify the expected time and location of initial band formation.
5) During the event
The existence of a banded feature allows for direct correlation of short-range model analyses and forecasts with radar, sounding, satellite, wind profiler, and other observations. Of particular interest is the correlation of the observed band with midlevel frontogenesis and weak moist symmetric stability. Based on the forecast frontogenesis and moist symmetric stability evolutions, detailed predictions of band movement, intensity, and dissipation can be included within Short Term Forecasts for specific counties or even portions of counties.
c. Summary
The proposed forecast strategy is summarized in Table 2. The strategy guides forecasters through the assessment of frontogenesis, weak moist symmetric stability, and moisture on a time- and scale-dependent basis. These ingredients are considered at all forecast projections, but the context (column 2 in Table 2) and scale (column 3 in Table 2) in which they are assessed change through time. As the banding event nears, the outlined threat area narrows from the scale of states, to counties, to portions of counties. Concurrent with this spatial focusing is an increasing specificity in the timing of banding as confidence increases for shorter-range forecast projections. Such information can be communicated through operational forecast products to first highlight the potential, then likelihood, and finally imminent occurrence of banding (Table 2).
4. Case example
a. Background
During December 2002, training outlining this mesoscale band forecast strategy was presented to NOAA/NWS forecast offices (Novak et al. 2005), of which the Albany, New York (ALY), and Binghamton, New York (BGM), Weather Forecast Offices (WFOs) were participants. A Weather Event Simulator (WES) practice case was also developed and released in conjunction with the training. WES is a software package that simulates the NOAA/NWS Advanced Weather Interactive Processing System (AWIPS) software environment, allowing forecasters to examine archived cases in simulated real time, analogous to flight simulators used in pilot training (Magsig and Page 2003). Just a few weeks after the forecast strategy training was given, the ALY and BGM County Warning Areas (CWAs) were directly affected by a mesoscale snowband as a strong coastal storm developed off the northeast U.S. coast during the afternoon and evening of 25 December 2002 (Christmas Day). Storm total snowfall accumulations exceeding 61 cm (24 in.) were widespread across central and eastern New York (Fig. 3) due largely to the snowband, with snowfall rates of 13 cm h−1 (5 in. h−1) officially recorded as the band moved through Albany, New York. Application of the forecast strategy was evident in forecast products issued by the BGM and ALY WFOs, providing an opportunity to illustrate the strategy and to consider its benefits and practical challenges in relation to forecast operations.
b. Application of the forecast strategy
1) Greater than 48 h before the event
On 22 and 23 December 2002 confidence in a major cyclogenesis event occurring Christmas Day off the northeast U.S. coast was supported by individual NCEP operational model runs and the NCEP SREF system consensus. For example, the 0900 UTC 23 December 2002 NCEP SREF system 60-h forecast ensemble mean of the mean sea level pressure valid at 2100 UTC 25 December 2002 featured a cyclone off the eastern U.S. coast, with standard deviation values maximized to the northeast of the cyclone center (Fig. 4a). A 60-h forecast spaghetti plot of the 1000-hPa mean sea level isobar valid at the same time (Fig. 4b) shows that all individual NCEP SREF system ensemble members featured a cyclone off the eastern U.S. coast, although a few members placed the surface cyclone farther northeast of the ensemble mean position, consistent with the location of the standard deviation maximum noted in Fig. 4a. The foregoing 60-h NCEP SREF system products valid at 2100 UTC 25 December 2002 suggest that there was relative certainty in the presence of a cyclone along the coast at this time, although there was uncertainty in its position and intensity. The operational 1200 UTC 23 December 2002 Eta and GFS models were largely within the envelope of the NCEP SREF system member solutions (Fig. 4b), and also were forecasting a cyclone that was to move northeast parallel to the northeast U.S. coast.
In recognition of the potential for cyclogenesis and associated heavy snowfall, Winter Storm Outlooks were issued by the ALY and BGM WFOs for eastern New York and northeast Pennsylvania early on 23 December 2002 (Figs. 5a,b). The situational awareness for the possibility of banding with this system was evident in the 0807 UTC [0307 eastern standard time (EST)] 23 December 2002 Area Forecast Discussion (AFD) issued by BGM, which stated, “Once models settle down [we] will have to evaluate the frontal structure for embedded midlevel frontogenesis in the dendritic zone and slantwise instability to see if banding will be present. Places could then jump over a foot. . . . ”
2) 24–48 h before the event
As the event drew within 48 h, the band ingredients were assessed to determine whether banding was possible (as suggested by previous shifts). Consistent with prior NCEP SREF system guidance, both the 0000 UTC 24 December 2002 GFS and Eta Model runs were predicting the formation of a surface low off the New Jersey coast by 1800 UTC 25 December 2002, with evidence of coupled jets (more prominent in the Eta Model forecast; Figs. 6a,c). The forecast surface low positions in both models had trended closer to the coast relative to the NCEP SREF system members (cf. Figs. 6a,c and Fig. 4b). Both the GFS and Eta Model forecasts also showed the formation of a closed 700-hPa low and associated deformation and frontogenesis over Pennsylvania and New York (Figs. 6b,d). However, differences in the shape of the midlevel circulation resulted in differences in the location of the deformation maximum, with the GFS model depicting the deformation maximum northwest of the surface low (Fig. 6b) and the Eta Model placing this maximum north of the surface low (Fig. 6d). Consequently, the orientation of the associated frontogenesis maximum was along a northeast–southwest axis in the GFS model forecast and along an east–west axis in the Eta Model forecast. Although the details varied, the forecast synoptic flow environment and associated upper-level wind, midlevel deformation, and frontogenesis fields resembled conceptual models of favorable banding environments (cf. Figs. 1 and 6).
Forecasters' awareness of the favorable banding environment was highlighted in AFDs, as the possibility of intense mesoscale band formation was recognized. Focus on the occurrence of a cyclogenesis event and the locations of the primary midlevel deformation zone and frontogenesis maximum was evident in the 0813 UTC (0313 EST) 24 December 2002 BGM discussion (Fig. 7a). In recognition of the potential for heavy banded snowfall, Winter Storm Watches were issued for the ALY and BGM CWAs at 0815 UTC (0315 EST) and 0920 UTC (0420 EST) 24 December 2002, respectively (Figs. 7b,c). The watch language highlighted the potential for very heavy snowfall rates and snowfall accumulations of 1–2 ft.
Guidance from the NCEP SREF system initialized at 0900 UTC 24 December 2002 provided increased confidence in a heavy precipitation event, showing at least a 90% probability of 0.25 in. (Fig. 8a) and a 50% probability of 0.50 in. (Fig. 8b) of accumulated precipitation occurring over the ALY and eastern BGM CWAs during the 6-h period ending at 0000 UTC 26 December 2002. As data from the 1200 UTC 24 December 2002 operational Eta and GFS model runs (not shown) continued to depict strong deformation and frontogenesis with small negative EPVg* values collocated over eastern Pennsylvania and New York during Christmas Day, forecasters recognized the increasing likelihood of heavy snowfall (Figs. 9a–c). Winter Storm Warnings for Christmas Day were issued by the BGM and ALY WFOs at 2031 UTC (1531 EST) and 2040 UTC (1540 EST) 24 December 2002, respectively (Figs. 9c,e). These warnings cited heavy snowfall rates and total accumulations of 1–2 ft, based largely on the anticipation of banding (Figs. 9b,d). It should be noted that high confidence in the occurrence of warning criteria snowfall in this case allowed the issuance of the warning with nearly 24-h lead time; typically, the predictability of mesoscale banding would limit the warning issuance to the 12–24-h period as noted in section 3b(3). Although forecasters expressed confidence in the occurrence of a major storm and outlined the general region of possible band formation, specifying the explicit location of the band was recognized as a remaining forecast question to be addressed closer to the event (Figs. 9a,b).
3) 12–24 h before the event
As the event drew within 24 h, forecasters assessed when and where banding was favored by further evaluating the frontogenesis, moist symmetric stability, and moisture fields. The 18-h forecast fields from the 0000 UTC 25 December 2002 GFS (Fig. 10) and Eta (Fig. 11) models were similar with respect to the depth and location of the 700-hPa low and associated location and orientation of the deformation (not shown) and frontogenesis maxima, stretching from eastern Pennsylvania through eastern New York. However, the different numerical formulations [e.g., limited-area grid point (Eta) versus global spectral (GFS)] and higher horizontal and vertical resolution of the Eta Model as compared to the GFS model allowed better simulation of mesoscale processes in the Eta Model. A comparison of the 80-km display grids from the GFS (GFS80; Figs. 10a,b) and Eta (Eta80; Figs. 11a,b) shows that the Eta forecasted stronger frontogenesis. The frontogenesis maximum forecast by the Eta was also ∼100–200 km farther north and west than the frontogenesis maximum forecast by the GFS.
The impact of model display resolution is evident by comparing Figs. 11a and 11c, as the frontogenesis maximum magnitude calculated on the 40-km display grid from the Eta (Eta40) is nearly three times the magnitude of the Eta80 maximum and approximately half the width. This increase in magnitude and decrease in scale of the frontogenesis maximum appears to be a result of the Eta40 resolving a sharp 700-hPa trough extending northeast from the 700-hPa low center. Additionally, the higher spatial and vertical resolution of the Eta40 better depicts the weakness in the midlevel EPVg* field coincident with the midlevel frontogenesis.
Cross sections through the GFS80 and Eta80 frontogenesis maxima (Figs. 10b and 11b) exhibited similar features as shown in conceptual models of banding environments (e.g., Fig. 2), including a region of weak conditional stability (and implied weak moist symmetric stability) above the 700-hPa frontogenesis maximum (note the steep slope of the saturation equivalent potential temperature contours) and associated narrow axis of ascent. However, the Eta80 frontogenesis maximum was stronger and farther northwest. The same cross section calculated using the Eta40 grids better depicts the weakness in the conditional stability, as the saturation equivalent potential temperature contours are nearly vertical above the 700-hPa frontogenesis maximum (Fig. 11d). Accordingly, the ascent in the Eta40 cross section exceeds −44 × 10−3 hPa s−1, which is nearly twice that shown in the GFS80 and Eta80 cross sections. This comparison highlights the importance of considering the resolution of display grids when assessing the band ingredients.
As discussed in section 3b(3), the calculation of EPV* has been shown to be quite sensitive to the choice of representative wind (geostrophic or full wind). To illustrate this difference, EPV* was calculated with the geostrophic and full wind using the Eta40 18-h forecast grids (Fig. 12). Inspection reveals that a large portion of the negative EPVg* field evident in Fig. 11c corresponds to an area of negative geostrophic absolute vorticity (Fig. 12a). This area of negative geostrophic absolute vorticity was found along the northeast–southwest axis of a short-wave ridge extending through eastern and central New York. Calculation of EPVf* (Fig. 12b) results in an area of small positive values corresponding to the area of negative values of EPVg* (Fig. 12a), but 50–100 km farther northwest and coincident with small positive values of the full-wind absolute vorticity. Jurewicz and Evans (2004) have similarly noted that the EPVg* fields in their analysis of the 6 January 2001 snowstorm over the northeast United States were negative due to negative values of geostrophic absolute vorticity.
Model QPF fields were also referenced by forecasters during the event to aid in the interpretation of the model diagnostic fields, confirm the likelihood of banding, and refine the predicted location of the band. The forecast 6-h accumulated precipitation ending at 0000 UTC 26 December 2002 from the 0000 UTC 25 December 2002 GFS and Eta models showed marked differences between the GFS80 (Fig. 13a) and Eta80 (Fig. 13b), with the GFS80 depicting a broad maximum of 24–36 mm (0.94–1.42 in.) along the coast while the Eta80 showed an elongated maximum exceeding 28 mm (1.10 in.) extending from northeast Pennsylvania through eastern New York into southwest Vermont. The Eta40 precipitation field better accentuates the forecast band, with an elongated 32-mm (1.26 in.) maximum over eastern New York (Fig. 13c). The Eta40 precipitation field was similar to forecast hourly precipitation fields from the Stony Brook University 12-km MM5 (Colle et al. 2003), which clearly resolved a band in the 1-h accumulated precipitation ending at 2100 UTC 25 December 2002 (Fig. 13d).
The early Christmas morning BGM AFD (Fig. 14) shows that given the presence of the band ingredients in both the GFS and Eta models, forecasters favored the precipitation solution from the Eta Model in recognition that the higher-resolution Eta would have greater skill in capturing the magnitudes and spatial scales of the frontogenesis maximum, EPVg* minimum, and ascent maximum. The model agreement in the wind, geopotential height, and temperature fields, coupled with the ascent and precipitation fields, gave forecasters confidence to place the likely area of banded snow in sections of eastern New York. The forecast band and sharp snowfall gradient to its north and west were integrated into experimental forecast products, including the National Digital Forecast Database (NDFD; Glahn and Ruth 2003) and BGM WFO Web page graphics posted at 0900 UTC (0400 EST) for the 12-h period ending at 2300 UTC (1800 EST) 25 December 2002 (Fig. 15a). Although maximum snowfall amounts were underpredicted, the location of the forecast snowfall accumulation maximum and the occurrence of a tight gradient in snowfall to the north and west of this maximum were well predicted (cf. Figs. 15a and 15b).
4) 0–12 h before the event
As the event drew within 12 h, forecasters assessed the RUC and Eta Model analyses and short-range forecasts in concert with observations to answer the critical question of precisely when and where band formation would occur. At 1500 UTC 25 December 2002, an area of intensifying precipitation over eastern Pennsylvania was correlated with a developing region of 700-hPa frontogenesis (Fig. 16a) and small negative EPVg* values (not shown). Although bright banding due to precipitation melting aloft enhanced reflectivity over eastern Pennsylvania, heavy precipitation was recorded in this region at this time. The 700-hPa frontogenesis maximum was forecast to intensify, narrow, and move northward into eastern New York by 1800 UTC 25 December 2002 (Fig. 16b), then slowly shift eastward and dissipate through 0000 UTC 26 December 2002 (Figs. 16c,d). The forecast midlevel EPVg* field exhibited small negative values between 1800 and 2100 UTC 25 December 2002 (Figs. 16b,c), with small positive values largely restored by 0000 UTC 26 December 2002 (Fig. 16d). AFDs issued by the ALY and BGM WFOs at 1540 UTC (1040 EST) and 1550 UTC (1050 EST) 25 December 2002, respectively (Figs. 17a,b), highlighted forecasters' assessment of the band ingredients. This assessment, in the context of continued run-to-run model consistency and the temporal proximity to the event, gave forecasters confidence to call for the “likelihood of very intense banding over the next 6 hours” approximately 3 h before initial band development (Fig. 17b).
5) During the event
Single-band formation occurred just after 1900 UTC (1400 EST) across eastern New York. As the event progressed through the afternoon, forecasters recognized that the band would persist. Short Term Forecasts for affected portions of the ALY CWA noted the extreme nature of the event including the possibility of more than a foot of snowfall accumulation during a 3-h period within the band (Fig. 18a). Subsequent Short Term Forecasts were issued as the band shifted southeast (Fig. 18b). As the 1800 UTC RUC forecast (not shown) predicted the weakening of midlevel frontogenesis and the restoration of small positive EPVg* values by 0000 UTC 26 December 2002, the final Short Term Forecasts associated with the band reduced the forecast hourly snowfall rates and predicted the demise of the band by 0330 UTC 26 December 2002 (2230 EST 25 December 2002; Fig. 18c). Regional radar and RUC 700-hPa frontogenesis analyses show that the observed band and frontogenesis coincide throughout the evolution of the event, with the frontogenesis correlated with the band intensity during its initial development at 1900 UTC (Fig. 19a), strengthening between 2100 and 2300 UTC 25 December 2002 (Figs. 19b,c), and weakening by 0100 UTC 26 December 2002 (Fig. 19d).
5. Discussion and summary
A forecast strategy to anticipate cold season mesoscale band formation within eastern U.S. cyclones was proposed and illustrated. This strategy draws on the assessment of three band ingredients—frontogenesis, weak moist symmetric stability, and moisture—in a time- and scale-dependent approach. At forecast projections greater than 48 h, the strategy focuses on assessing the possibility of cyclogenesis to alert the forecaster that an environment supporting band formation may be present. At forecast projections of 24–48 h, the band ingredients are assessed in the context of the broader synoptic flow to determine whether banding is possible. At forecast projections within 12–24 h of the event, the strategy focuses on assessment of the band ingredients to outline when and where band formation is favored. Within 12 h of the event, the forecast ingredients are assessed and correlated with real-time observations, from which specific band predictions can be made.
Application of the strategy to the 25 December 2002 snowstorm showed how the forecast strategy can enhance situational awareness and lead to accurate forecasts of band development and evolution. At forecast projections greater than 48 h, the expectation of a cyclogenesis event prompted forecasters to recognize the potential for the development of an environment conducive to band formation and to focus subsequent attention on the assessment of the band ingredients. At forecast projections of 24–48 h, plan-view and cross-sectional analyses of gridded model fields in conjunction with high-resolution model guidance were used to assess the band ingredients to determine the likelihood of banding and to outline the threat area. As the event neared, short-range model forecasts coupled with real-time observations were used to refine the threat area, monitor band formation, and accurately forecast band movement, intensity, and dissipation. The issuance of a Short Term Forecast calling for a foot of snowfall in a 3-h period is particularly noteworthy. It seems unlikely that such an extreme short-term snowfall forecast would have been issued in the absence of prior assessment and situational awareness. This example shows that accurate operational forecasts of mesoscale bands can be made based on current conceptual understanding, observational datasets, and modeling capabilities.
Application of the forecast strategy to the 25 December 2002 snowstorm also illustrates a number of operational challenges in anticipating band formation. For example, quantitative assessment of the ingredients can be influenced by both the model and display resolution, which affects the magnitudes and scales of narrow, elongated patterns of frontogenesis, EPV*, vertical motion, and accumulated precipitation indicative of band formation. This case also highlights the question of which form of EPV* (EPVg* or EPVf*) is of greater operational value. This diagnostic revealed the presence of an area of negative EPVg* values, and a corresponding area of small positive EPVf* values, above the frontogenesis maximum. Consistent with the findings of Novak et al. (2004), the proposed strategy favors the interpretation that small positive values of EPVf* signify weak moist symmetric stability. This interpretation is favored since the full wind is likely to be more representative than the geostrophic wind in evaluating moist symmetric stability in the curved flow environments of bands in eastern U.S. cyclones. Regardless of which form of EPV* is adopted for diagnosis in the forecast strategy (EPVg* or EPVf*), weak moist symmetric stability is manifested as a local minimum in the respective EPV* fields. Nevertheless, differences in location of the respective EPV* minima on the order of 50–100 km were documented for the 25 December 2002 case, which are significant in comparison to the width of the snowband.
Since the 25 December 2002 case exhibited intense band formation, the question arises whether the forecast strategy may be extended to cases exhibiting weaker bands (i.e., bands with smaller dBZ values and snowfall rates than the 25 December 2002 band). Recent case examples allow consideration of this question. For example, Vallee and Vallier-Talbot (2004) presented a case in which the Taunton, Massachusetts (BOX), WFO used the forecast strategy to accurately adjust snowfall amounts for a banded snowfall event that exhibited snowfall rates on the order of 2 in. h−1 [this event corresponds to case 3 discussed by Banacos (2003)]. There are also elements of the forecast strategy presented in Schumacher (2003), where examples of successful band forecasts are shown for the Sioux Falls, South Dakota (FSD), WFO CWA, including banded events weaker than in the 25 December 2002 case. In a comparison of two banded snowfall events in central New York, Jurewicz and Evans (2004) showed that the weaker banded event exhibited the band ingredients (i.e., frontogenesis, weak moist symmetric stability, and moisture), but the vertical extent of the moisture was limited and the frontogenesis was slightly weaker and shallower than in their stronger case. These collective examples suggest that the forecast strategy can be applied successfully to weaker banded events, subject to the caveat that the band ingredients may occur in a less favorable combination for weaker events.
It is anticipated that the predictability of the band ingredients and associated mesoscale banding will be strongly case dependent. Operational experience suggests that the strategy may be applied successfully to cases that are less predictable than the 25 December 2002 case, although the forecast projections to which the respective forecast strategy questions apply may be shorter than those specified in Table 2. For example, assessment of ensemble and deterministic model guidance 12–24 h in advance of a given event may show that considerable uncertainty remains concerning the location, as well as the intensity, of deformation and related frontogenesis associated with a deepening cyclone. Consequently, forecasters may not have sufficient confidence to answer the “when” and “where” questions regarding the potential for band formation at this forecast projection, so assessment would continue. A key benefit of this time- and scale-dependent forecast strategy is that it enhances forecaster situational awareness. In cases of reduced predictability, such enhanced situational awareness may facilitate the rapid updating of forecasts as confidence increases in the likelihood of band formation and in the details of the subsequent band evolution. Additional case examples would serve to clarify the benefits and practical challenges associated with the forecast strategy in cases that are less predictable than that of 25 December 2002.
Although ensemble guidance can aid in assessing the uncertainty of forecasts of the synoptic and mesoscale flow evolution and in establishing a level of confidence in model forecasts of heavy precipitation, the resolution of contemporary operational ensemble forecast systems limits their utility in diagnosing the specific band ingredients. The advent of high-resolution SREF systems, such as the Stony Brook University MM5 ensemble system (Colle et al. 2004), provides the opportunity to assess the predictability of the band ingredients. For example, one might use high-resolution short-range ensemble forecast systems to depict an envelope of solutions concerning the locations and magnitudes of frontogenesis, moist symmetric stability, and moisture that signify the likelihood of band formation. Future investigations are planned to explore the use of high-resolution SREF systems in forecasting mesoscale bands.
Acknowledgments
The authors wish to thank Peter Banacos (NOAA/NWS/SPC), Brian Colle (Stony Brook University), Michael Evans (NOAA/NWS, Binghamton, New York), Heather Hauser (NOAA/NWS Eastern Region Headquarters), Dave Nicosia (NOAA/NWS, Binghamton, New York), David Schultz (NSSL/CIMMS), Phil Schumacher (NOAA/NWS, Sioux Falls, South Dakota), and Warren Snyder (NOAA/NWS, Albany, New York) for insightful comments and discussions concerning this work. Dolores Kiessling (COMET) and Bill Bua (COMET/NOAA/NWS) aided in acquiring data necessary for this study. Bill Bua also aided in the preparation of Figs. 4, 8, and 13. Two anonymous reviewers provided constructive comments leading to improvements in the presentation of this work. RUC data were obtained from the Atmospheric Radiation Measurement Program. This work was supported in part by NOAA Grant NA07WA0458, awarded to the University at Albany, SUNY, as part of the Collaborative Science, Technology, and Applied Research program.
REFERENCES
Banacos, P. C., 2003: Short-range prediction of banded precipitation associated with deformation and frontogenetic forcing. Preprints, 10th Conf. on Mesoscale Processes, Portland, OR, Amer. Meteor. Soc., CD-ROM, P1.7.
Benjamin, S. G., Grell G. A. , Brown J. M. , Smirnova T. G. , and Bleck R. , 2004: Mesoscale weather prediction with the RUC hybrid isentropic-terrain-following coordinate model. Mon. Wea. Rev., 132 , 473–494.
Black, T. L., 1994: The new NMC mesoscale Eta model: Description and forecast examples. Wea. Forecasting, 9 , 265–278.
Clark, J. H. E., James R. P. , and Grumm R. H. , 2002: A reexamination of the mechanisms responsible for banded precipitation. Mon. Wea. Rev., 130 , 3074–3086.
Colle, B. A., Olson J. B. , and Tongue J. S. , 2003: Multiseason verification of the MM5. Part I: Comparison with the Eta Model over the central and eastern United States and impact of MM5 resolution. Wea. Forecasting, 18 , 431–457.
Colle, B. A., Jones M. , Tongue J. S. , and Olson J. B. , 2004: An investigation of the mesoscale predictability over the northeast U.S. Preprints, 20th Conf. on Weather Analysis and Forecasting, Seattle, WA, Amer. Meteor. Soc., CD-ROM, 23.2.
Dalcher, A., and Kalnay E. , 1987: Error growth and predictability in operational ECMWF forecasts. Tellus, 39A , 474–491.
desJardins, M. L., Brill K. F. , and Schotz S. S. , 1991: Use of GEMPAK on UNIX workstations. Proc. Seventh Int. Conference on Interactive Information and Processing Systems for Meteorology, Oceanography, and Hydrology, New Orleans, LA, Amer. Meteor. Soc., 449–453.
Doswell C. A. III, , Brooks H. E. , and Maddox R. A. , 1996: Flash flood forecasting: An ingredients-based methodology. Wea. Forecasting, 11 , 560–581.
Droegemeier, K. K., 1997: The numerical prediction of thunderstorms: Challenges, potential benefits, and results from real-time operational tests. WMO Bulletin, Vol. 46, World Meteorological Organization, Geneva, Switzerland, 324–336.
Du, J., and Tracton M. S. , 2001: Implementation of a real-time short-range ensemble forecasting system at NCEP: An update. Preprints, Ninth Conf. on Mesoscale Processes, Fort Lauderdale, FL, Amer. Meteor. Soc., 355–356.
Dudhia, J., 1993: A nonhydrostatic version of the Penn State–NCAR mesoscale model: Validation tests and simulation of an Atlantic cyclone and cold front. Mon. Wea. Rev., 121 , 1493–1513.
Environmental Modeling Center, cited. 2005: The GFS atmospheric model. NCEP Office Note 442, 14 pp. [Available online at http://www.emc.ncep.noaa.gov/officenotes/newernotes/on442.pdf.].
Glahn, H. R., and Ruth D. P. , 2003: The new digital forecast database of the National Weather Service. Bull. Amer. Meteor. Soc., 84 , 195–201.
Janjić, Z. I., 2004: The NCEP WRF core. Preprints, 20th Conf. on Weather Analysis and Forecasting, Seattle, WA, Amer. Meteor. Soc., CD-ROM, 12.7.
Jurewicz, M. L., and Evans M. S. , 2004: A comparison of two banded heavy snowstorms with very different synoptic settings. Wea. Forecasting, 19 , 1011–1028.
Klazura, G. E., and Imy D. A. , 1993: A description of the initial set of analysis products available from the NEXRAD WSR-88D system. Bull. Amer. Meteor. Soc., 74 , 1293–1311.
Magsig, M. A., and Page E. M. , 2003: Weather event simulator implementation and future development. Preprints, 20th Int. Conf. on Interactive Information Processing Systems for Meteorology, Oceanography, and Hydrology, Long Beach, CA, Amer. Meteor. Soc., CD-ROM, 12.9.
Martin, J. E., 1998a: The structure and evolution of a continental winter cyclone. Part I: Frontal structure and the occlusion process. Mon. Wea. Rev., 126 , 303–328.
Martin, J. E., 1998b: The structure and evolution of a continental winter cyclone. Part II: Frontal forcing of an extreme snow event. Mon. Wea. Rev., 126 , 329–348.
McNulty, R. P., 1978: On upper tropospheric kinematics and severe weather occurrence. Mon. Wea. Rev., 106 , 662–672.
Miller, J. E., 1948: On the concept of frontogenesis. J. Meteor., 5 , 169–171.
Moninger, W. R., Mamrosh R. D. , and Pauley P. M. , 2003: Automated meteorological reports from commercial aircraft. Bull. Amer. Meteor. Soc., 84 , 203–216.
Moore, J. T., Graves C. E. , Ng S. , and Smith J. L. , 2005: A process-oriented methodology toward understanding the organization of an extensive mesoscale snowband: A diagnostic case study of 4–5 December 1999. Wea. Forecasting, 20 , 35–50.
Nicosia, D. J., and Grumm R. H. , 1999: Mesoscale band formation in three major northeastern United States snowstorms. Wea. Forecasting, 14 , 346–368.
Novak, D. R., Bosart L. F. , Keyser D. , and Waldstreicher J. S. , 2004: An observational study of cold season–banded precipitation in northeast U.S. cyclones. Wea. Forecasting, 19 , 993–1010.
Novak, D. R., Waldstreicher J. S. , Bosart L. F. , and Keyser D. , cited. 2005: Anticipating mesoscale band formation in winter storms. Webcast. Virtual Institute for Satellite Integration Training. [Available online at http://www.cira.colostate.edu/ramm/visit/banding.html.].
Roebber, P. J., Schultz D. M. , and Romero R. , 2002: Synoptic regulation of the 3 May 1999 tornado outbreak. Wea. Forecasting, 17 , 399–429.
Roebber, P. J., Bruening S. L. , Schultz D. M. , and Cortinas J. V. Jr., 2003: Improving snowfall forecasting by diagnosing snow density. Wea. Forecasting, 18 , 264–287.
Roebber, P. J., Schultz D. M. , Colle B. A. , and Stensrud D. J. , 2004: Toward improved prediction: High-resolution and ensemble modeling systems in operations. Wea. Forecasting, 19 , 936–949.
Rogers, E., Black T. , Ferrier B. , Lin Y. , Parrish D. , and DiMego G. , 2001: Changes to the NCEP Meso Eta Analysis and Forecast System: Increase in resolution, new cloud microphysics, modified precipitation assimilation, modified 3DVAR analysis. NOAA/NWS Technical Procedures Bulletin 488, 21 pp. [Available online at http://www.emc.ncep.noaa.gov/mmb/mmbpll/eta12tpb/; also available from Office of Climate, Water, and Weather Services, National Weather Service, 1325 East–West Highway, Silver Spring, MD 20910.].
Sanders, F., and Bosart L. F. , 1985: Mesoscale structure in the megalopolitan snowstorm of 11–12 February 1983. Part I: Frontogenetical forcing and symmetric instability. J. Atmos. Sci., 42 , 1050–1061.
Schultz, D. M., and Schumacher P. N. , 1999: The use and misuse of conditional symmetric instability. Mon. Wea. Rev., 127 , 2709–2732. Corrigendum. 128 , 1573.
Schultz, D. M., Keyser D. , and Bosart L. F. , 1998: The effect of large-scale flow on low-level frontal structure and evolution in midlatitude cyclones. Mon. Wea. Rev., 126 , 1767–1791.
Schumacher, P. N., 2003: An example of forecasting mesoscale bands in an operational environment. Preprints, 10th Conf. on Mesoscale Processes, Portland, OR, Amer. Meteor. Soc, CD-ROM, P 1.11.
Sivillo, J. K., Ahlquist J. E. , and Toth Z. , 1997: An ensemble forecasting primer. Wea. Forecasting, 12 , 809–818.
Snellman, L. W., 1982: Impact of AFOS on operational forecasting. Preprints, Ninth Conf. on Weather Forecasting and Analysis, Seattle, WA, Amer. Meteor. Soc., 13–15.
Snellman, L. W., and Thaler E. , 1993: Forecast process. COMET, CD-ROM version 1.0. [Available from COMET, UCAR, P.O. Box 3000, Boulder, CO 80307.].
Stensrud, D. J., Brooks H. E. , Du J. , Tracton M. S. , and Rogers E. , 1999: Using ensembles for short-range forecasting. Mon. Wea. Rev., 127 , 433–446.
Thorpe, A. J., and Emanuel K. A. , 1985: Frontogenesis in the presence of small stability to slantwise convection. J. Atmos. Sci., 42 , 1809–1824.
Toth, Z., Kalnay E. , Tracton S. M. , Wobus R. , and Irwin J. , 1997: A synoptic evaluation of the NCEP ensemble. Wea. Forecasting, 12 , 140–153.
Toth, Z., Zhu Y. , and Marchok T. , 2001: The use of ensembles to identify forecasts with small and large uncertainty. Wea. Forecasting, 16 , 463–477.
Trapp, R. J., Schultz D. M. , Ryzhkov A. V. , and Holle R. L. , 2001: Multiscale structure and evolution of an Oklahoma winter precipitation event. Mon. Wea. Rev., 129 , 486–501.
Vallee, D. R., and Vallier-Talbot E. , 2004: Snowin' to beat the band: Using satellite imagery and Local Analysis and Prediction System (LAPS) output to diagnose the rapid development of a mesoscale snow band. Presentations of the 20th Conference on Weather Analysis and Forecasting, Seattle, WA, Amer. Meteor. Soc., Presentation 2.1.
Wetzel, S. W., and Martin J. E. , 2001: An operational ingredients-based methodology for forecasting midlatitude winter season precipitation. Wea. Forecasting, 16 , 156–167.
Wiesmueller, J. L., and Zubrick S. M. , 1998: Evaluation and application of conditional symmetric instability, equivalent potential vorticity, and frontogenetical forcing in an operational forecasting environment. Wea. Forecasting, 13 , 84–101.
Zhang, F., Snyder C. , and Rotunno R. , 2003: Effects of moist convection on mesoscale predictability. J. Atmos. Sci., 60 , 1173–1185.
Conceptual models of synoptic and mesoscale environments conducive to band formation from (a) Nicosia and Grumm (1999) and (b) Novak et al. (2004). Features shown in (a) are labeled. Features shown in (b) include midlevel frontogenesis (shading), midlevel deformation zone (encompassed by scalloped line) and associated primary dilatation axes (dashed lines), midlevel streamlines (solid lines), and upper-level jet cores (wide dashed arrows).
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
Schematic cross section of the environment for a banded frontal zone. Saturation equivalent potential temperature (dashed), frontogenesis (ellipse), and transverse circulation (arrows) are shown, with dry air intrusion (light shading; an X depicts flow into the plane of the cross section) and areas exhibiting weak moist symmetric stability (WMSS; dark shading) overlaid. Expected location of precipitation band indicated by an asterisk (*) along the x axis. [Adapted from Moore et al. (2005, Fig. 15b)]
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
The 25–26 Dec 2002 storm total snowfall (shaded; in.), with the BGM and ALY WFO CWAs outlined in thick solid lines.
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
(a) The 60-h forecast ensemble mean (solid, contoured every 4 hPa) and standard deviation (shaded according to scale starting at 1 hPa) of the mean sea level pressure from the 0900 UTC 23 Dec 2002 NCEP SREF system valid at 2100 UTC 25 Dec 2002. (b) The 60-h forecast spaghetti plot of the 1000-hPa mean sea level isobar from the 0900 UTC 23 Dec 2002 NCEP SREF system valid at 2100 UTC 25 Dec 2002, with individual ensemble members (thin gray) and ensemble mean (black solid) shown. The 57-h forecast 1000-hPa isobars from the 1200 UTC 23 Dec 2002 operational GFS (thick gray; labeled G) and Eta (thick gray; labeled E) models valid at 2100 UTC 25 Dec 2002 are also shown.
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
(a) Excerpts from the Winter Storm Outlook issued by the ALY WFO at 0825 UTC (0325 EST) 23 Dec 2002. (b) Excerpts from the Winter Storm Outlook issued by the BGM WFO at 0900 UTC (0400 EST) 23 Dec 2002.
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
(a) GFS80 42-h forecast mean sea level pressure (solid, contoured every 4 hPa) and 300-hPa wind speed (shaded according to scale starting at 50 m s−1) valid at 1800 UTC 25 Dec 2002. (b) Same as in (a), except for 700-hPa geopotential height (solid, contoured every 3 dam), 700-hPa resultant deformation equal to 8 × 10−5 s−1 (dotted), and 700-hPa Miller 2D frontogenesis [positive values shaded according to scale starting at 0.5°C (100 km)−1 (3 h)−1]. (c) Same as in (a) except for Eta80 forecast. (d) Same as in (b) except for Eta80 forecast.
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
(a) Excerpts from the AFD issued by the BGM WFO at 0813 UTC (0313 EST) 24 Dec 2002. (b) Excerpts from the Winter Storm Watch issued by the ALY WFO at 0815 UTC (0315 EST) 24 Dec 2002. (c) Same as in (b) except product issued by the BGM WFO at 0920 UTC (0420 EST) 24 Dec 2002.
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
Probability (contoured every 20% starting at 10%) of accumulated precipitation exceeding (a) 0.25 and (b) 0.50 in. during the 6-h period ending with the 39-h forecast from the 0900 UTC 24 Dec 2002 NCEP SREF system valid at 0000 UTC 26 Dec 2002.
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
(a) Excerpts from the AFD issued by the BGM WFO at 1535 UTC (1035 EST) 24 Dec 2002. (b) Same as in (a) except product issued at 2025 UTC (1525 EST) 24 Dec 2002. (c) Excerpts from the Winter Storm Warning issued by the BGM WFO at 2031 UTC (1531 EST) 24 Dec 2002. (d) Excerpt from the AFD issued by the ALY WFO at 2035 UTC (1535 EST) 24 Dec 2002. (e) Same as in (c) except product issued by the ALY WFO at 2040 UTC (1540 EST) 24 Dec 2002.
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
(a) GFS80 18-h forecast 700-hPa geopotential height (thick solid, contoured every 3 dam), geostrophic saturation equivalent potential vorticity calculated over the 700–600-hPa layer [thin solid, contoured every 0.5 PVU at and below 0 PVU (1 PVU = 10−6 m2 s−1 K kg−1)], and 700-hPa Miller 2D frontogenesis [positive values shaded according to scale starting at 1.0°C (100 km)−1 (3 h)−1] valid at 1800 UTC 25 Dec 2002. (b) Cross section through midlevel frontogenesis maximum [orientation given in (a)] showing saturation equivalent potential temperature (solid, contoured every 3 K), Miller 2D frontogenesis [positive values shaded according to scale starting at 1.0°C (100 km)−1 (3 h)−1], and vertical velocity (dashed, contoured every 4 × 10−3 hPa s−1 at and below −4 × 10−3 hPa s−1).
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
(a) Same as in Fig. 10a except for Eta80 forecast. (b) Same as in Fig. 10b except for Eta80 forecast. (c) Same as in (a) except for Eta40 forecast. (d) Same as in (b) except for Eta40 forecast
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
(a) Eta40 18-h forecast 700-hPa geopotential height (thick solid, contoured every 3 dam), geostrophic saturation equivalent potential vorticity calculated over the 700–600-hPa layer (thin solid, contoured every 0.5 PVU at and below 0 PVU), and 700-hPa geostrophic absolute vorticity (negative values shaded according to scale in 10−5 s−1) valid at 1800 UTC 25 Dec 2002. (b) Same as in (a) except saturation equivalent potential vorticity calculated using the full wind (thin, solid, contoured at 0.1, 0.2, and 0.4 PVU), and absolute vorticity calculated using the full wind.
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
Six (6)-h accumulated precipitation (shaded according to scale in mm) ending at 0000 UTC 26 Dec 2002 from the (a) GFS80, (b) Eta80, and (c) Eta40 forecasts initialized at 0000 UTC 25 Dec 2002. (d) Same as in (a), except 1-h accumulated precipitation ending at 2100 UTC 25 Dec 2002 from the Stony Brook University 12 km MM5.
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
Excerpt from the AFD issued by the BGM WFO at 0807 UTC (0307 EST) 25 Dec 2002.
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
(a) BGM WFO forecast snowfall accumulation (shaded, in.) for the 12-h period 1100–2300 UTC (0600–1800 EST) 25 Dec 2002. (b) Storm total snowfall (shaded, in.) across the BGM CWA.
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
(a) WSR-88D radar mosaic (shaded according to scale starting at 5 dBZ) at 1500 UTC 25 Dec 2002 with the 1500 UTC RUC40 analysis 700-hPa Miller 2D frontogenesis overlaid [black solid, contoured every 1.5°C (100 km)−1 (3 h)−1 starting at 1.5°C (100 km)−1 (3 h)−1]. (b) Eta40 6-h forecast 700-hPa geopotential height (solid, contoured every 3 dam), 700-hPa Miller 2D frontogenesis [positive values shaded according to scale starting at 1.5°C (100 km)−1 (3 h)−1], and geostrophic saturation equivalent potential vorticity calculated over the 700–600-hPa layer (thin solid, contoured every 0.5 PVU at and below 0 PVU) valid at 1800 UTC 25 Dec 2002. (c) Same as in (b) except for 9-h forecast valid at 2100 UTC 25 Dec 2002. (d) Same as in (b) except for 12-h forecast valid at 0000 UTC 26 Dec 2002.
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
Excerpts from the AFDs issued by (a) the ALY WFO at 1540 UTC (1040 EST) and (b) the BGM WFO at 1550 UTC (1050 EST) 25 Dec 2002.
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
(a) Excerpts from the Short Term Forecast issued by the ALY WFO at 2037 UTC (1537 EST) 25 Dec 2002, and corresponding 0.5° elevation angle WSR-88D radar imagery (shaded according to scale starting at 5 dBZ) near issuance time with Short Term Forecast area outlined in thick black. (b) Same as in (a) except for excerpts from the Short Term Forecast issued at 2330 UTC (1830 EST) 25 Dec 2002. (c) Same as in (a) except for excerpts from the Short Term Forecast issued at 0212 UTC 26 Dec 2002 (2112 EST 25 Dec 2002).
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
(a) WSR-88D radar mosaic (shaded according to scale starting at 5 dBZ) at 1900 UTC 25 Dec 2002 with the 1900 UTC RUC40 analysis 700-hPa Miller 2D frontogenesis overlaid [black solid, contoured every 1.5°C (100 km)−1 (3 h)−1 starting at 1.5°C (100 km)−1 (3 h)−1]. (b) Same as in (a) except for 2100 UTC 25 Dec 2002. (c) Same as in (a) except for 2300 UTC 25 Dec 2002. (d) Same as in (a) except for 0100 UTC 26 Dec 2002.
Citation: Weather and Forecasting 21, 1; 10.1175/WAF907.1
NOAA/NWS product definitions. Note that snow (or ice) accumulation criteria for watches/warnings vary by forecast location.
Forecast strategy for anticipating mesoscale band formation within eastern U.S. cyclones as relating to assessment of band ingredients (frontogenesis, weak moist symmetric stability, and moisture).
The Eta Model has been renamed the North American Mesoscale (NAM) model as of 25 January 2005. Documentation of this name change can be found online (http://www.nws.noaa.gov/om/notification/notif04/tin04-58avn_eta_rename_aaa.txt).
