Abstract
In 2012, an observing-systems component was added to the activities of the annual NOAA/Hazardous Weather Testbed Spring Forecasting Experiment (SFE). One of the systems tested is a low-cost radiosonde system from InterMet Systems, which is being used more in research applications as an alternative to the reliable, but costly, Vaisala RS92 system. The collection of soundings in pre-convective and near-storm environments will be an important component of upcoming field projects and money could be saved by using an InterMet system. However, little is known about how the soundings from the InterMet radiosondes compare to the Vaisala soundings. The first part of this talk will cover an intercomparison of these two sounding systems performed last spring in the HWT/SFE and show that the InterMet radiosonde profiles compare very favorably to the Vaisala profiles.

The second part of this talk covers an evaluation of forecasts over two spring seasons from five convection-allowing configurations of the WRF-ARW model that vary only by the turbulence parameterization scheme, including three "local" schemes (MYJ, QNSE, MYNN) and two schemes that include "non-local" mixing (YSU and ACM2). CAPS ran these forecasts as part of the suite of convection-allowing model forecasts provided for HWT/SFE activities in 2011 and 2012. The forecasts are compared to radiosonde observations upstream from deep convection to gain a better understanding of the thermodynamic characteristics of these schemes in this regime. The results show that the MYNN scheme alleviates the typical cool (warm), moist (dry) bias of local (non-local) schemes in afternoon/evening convective boundary layers upstream from convection, but other aspects of the forecasts are not improved, and are even degraded, with the use of MYNN.

Abstract
The primary focus of this talk will be on a series of numerical model forecasts made of a lake-effect snow event. These forecasts examine how dependent QPF is on the choice of microphysical parameterization. Notable differences are observed in the amount and distribution of precipitation with some schemes having very narrow and heavy precipitation shields and others having a more diffuse pattern. Causes for these different patterns will be discussed as well as some limitations to validating QPF in the lake-effect snow environment. The talk will then segue into other winter weather topics the author is involved with at NSSL. These include the development of a new winter surface hydrometeor classification algorithm, some forays into blizzard potential forecasting, and theoretical aspects of orographic precipitation.

Abstract
A severe thunderstorm wind gust climatology spanning 2003-2009 for the contiguous United States is developed using measured Automated Surface Observing System (ASOS) and Automated Weather Observing System (AWOS) wind gusts. Archived severe report information from the National Climatic Data Center’s Storm Data and single site volumetric radar data are used to identify severe wind gust observations [> 50 kt (25.7 m s-1)] associated with thunderstorms and to classify the convective mode of the storms.

The measured severe wind gust distribution, comprising only 2% of all severe gusts, is examined with respect to radar-based convective modes. The convective mode scheme presented herein focuses on three primary radar-based storm categories: supercell, quasi-linear convective systems (QLCS), and disorganized. Measured severe gust frequency revealed distinct spatial patterns, where the High Plains received the greatest number of gusts and occurred most often in the late spring and summer months. Severe wind gusts produced by supercells were most frequent over the Plains, while those from QLCS gusts were most frequent in the Plains and Midwest. Meanwhile, disorganized storms produced most of their severe gusts in the Plains and Intermountain West. A reverse spatial distribution signal exists in the location between the maximum measured severe wind gust corridor located over the High Plains and the maximum in all severe thunderstorm wind reports from Storm Data, located near and west of the southern Appalachians.

Abstract
The impacts of cloud microphysical processes on prediction of tropical cyclone environments, track, intensity, and structure are examined for two microphysical parameterizations using the Coupled Ocean / Atmosphere Mesoscale Prediction System ­Tropical Cyclone model. The control microphysical parameterization is a relatively typical single-moment scheme with five hydrometeor species: cloud water and ice, rain, snow, and graupel. An alternative newer method uses a hybrid approach of double-moment in cloud ice and rain and single moment in the other three species. Basin-scale synoptic flow simulations point to important differences between these two schemes. The upper level cloud ice concentrations produced by the control scheme are up to two orders of magnitude greater than by the newer scheme, primarily due to differing assumptions concerning the ice nuclei concentration. Significant (1-2ºC) warm biases near the 300-hPa level in the control experiments were not present using the newer scheme. The warm bias in the control simulations is associated with the longwave radiative heating near the base of the cloud ice layer.

The two schemes produced different track and intensity forecasts for 15 Atlantic storms. Rightward cross-track bias and positive intensity bias in the control forecasts are significantly reduced using the newer scheme. Synthetic satellite imagery of Hurricane Igor (2010) shows more realistic brightness temperatures from the simulations using the newer scheme, in which the inner core structure is clearly discernible. Applying the synthetic satellite imagery in both quantitative and qualitative analyses helped to pinpoint the issue of excessive upper-level cloud ice in the control scheme.

Abstract
During the second field phase of the Verifications of the Origins of Rotation in Tornadoes Experiment, Part 2 (VORTEX2), a slow moving supercell thunderstorm was observed by the mobile armada on 26 May 2010 just northeast of Denver, Colorado. In this study, observations from seven mobile radars; SR1, SR2, NOXP, DOW6, DOW7, TTUKa1 and TTUKa2 along with surface thermodynamic measurements form the fleet of mobile mesonets have been used to analyze the three dimensional wind field and thermodynamic structure of the 26 May storm for the period 2230 UTC to 2330 UTC.

Analyses from 2230 UTC to 2250 UTC reveal the cyclic nature of the storm with an occluded circulation exiting the rear of the storm simultaneously with the development of a new low-level mesocyclone along the leading edge of the hook echo. Pulses in the rear flank downdraft (RFD) and the occlusion downdraft were associated with an increase in the low-level vertical vorticity maxima within the main updraft where maximum stretching occurs. Yet, the high base storm was unable to generate a long-lived, strong, small-scale circulation associated with a tornado vortex. Instead, the storm did produce two short-lived, small-scale vortices. Thermodynamic measurements show the structure of the inflow and outflow regions of the storm as well as regions in the forward flank. The evolution of the RFD and occlusion downdraft was then compared to the evolution of two different storms sampled with the two mobile C-band radars (SR1 and SR2) on 29 May 2004 near Geary, OK and 19 June 2010 near Concordia, KS. Similarities and differences in the downdraft evolution of these three storms were investigated.

Abstract
The vortex detection and characterization method developed in Potvin et al. (2009, 2011) has been modified to operate on Cartesian horizontal wind analyses. Tests with high-resolution supercell simulations indicate the algorithm successfully detects intense vortices and estimates their size and maximum tangential winds, even in the presence of complex flow (Potvin 2013). The technique could therefore be very useful to investigating relationships between mesocyclone and tornado characteristics, and to detecting tornadoes, mesocyclones and mesovortices in real-time ensemble forecasts. An overview of the method and highlights of the tests with the supercell simulations will be presented.

The focus of the talk will be on preliminary efforts to develop a tornado detection algorithm based on the “vorticity line” method of Cai (2005). The underlying hypothesis of the vorticity line method is that vertical vorticity is scale-invariant [i.e., log(vorticity) vs. log(scale) is linear] within mesocyclones. This would allow vorticity at smaller (e.g., tornadic) scales to be reliably diagnosed from the vorticity at larger scales. While analyses of a 75 m tornadic supercell simulation and Doppler velocity observations of real tornadoes suggest mesocyclonic vorticity is not truly scale-invariant, they also support Cai’s conclusion that mesocyclones with steeper vorticity or pseudo-vorticity lines on radar-observed scales are more likely to be tornadic. These results and their implications for operational tornado detection will be discussed.

Abstract
Lightning, and in particular dry thunderstorms, are a major source of wildfire initiation each year. While several studies have been performed to predict cloud-to-ground (CG) lightning flashes, little to no work includes precipitation information with regards to dry thunder. We use the perfect prognosis technique over a 40 km Contiguous United States (CONUS) domain and 10km Alaska domain to try and predict different lightning and precipitation parameters. Development data comes from the 32km North American Regional Reanalysis (NARR) data set during the warm season (May-Sep) over 12 years (2000-2011). Principal Component Analysis helps use the most meaningful atmospheric parameters associated with dry thunderstorms for the predictive equations. Tests show that keeping 12 out of 140+ principal components explains much of variations in dry thunderstorm patterns. Predictive equations are created in Splus using a generalized linear model structure. We then feed GFS data into these equations to create forecasts over the two domains. Preliminary results show that the forecasts help identify broad regions of lightning/low precipitation potential, but have noticeable impacts from the GFS model biases and dependence on climatology. Other information, such as fuels and standard fire weather indices, could help to improve location accuracy of such forecasts.

Abstract
On 24 May, 2011, a severe weather outbreak spawned a series of strong to violent tornadoes across central Oklahoma. Data were collected from two of these tornadoes by a rapid-scanning, X-band, polarimetric mobile radar (RaXPol). The acquired dataset encompasses the intensification, mature and dissipation phases of an EF-3 tornado, and the genesis, intensification and mature phases of a subsequent EF-5 tornado. Volumetric observations over 360 degree PPIs at 9 elevation angles were collected on ~15 second time scales, and near-surface single elevation PPIs were collected every 2 seconds for a period of ~6 minutes during intensification of tornado 2. This study examines various aspects of the tornadoes’ evolutions in an attempt to address the following questions: 1) How does the rotation associated with the tornadic vortices evolve during tornadogenesis, intensification and decay? 2) How does the three-dimensional structure of the tornadoes change with time? 3) How are storm-scale features (such as gust fronts and downdrafts) involved with the decay and genesis processes? 4) How is the tornado debris signature evident in the polarimetric variables related to the wind field? and 5) What other features pertinent to tornado evolution can be observed from rapid-scan data? In order to answer these questions, single-Doppler analyses are performed that examine the evolution of rotation over all times and heights and that reconstruct the reflectivity, radial velocity, and cross-correlation coefficient parameters onto three-dimensional grids to examine the tornado structure and parameter interrelations. Additionally, a brief period of rapid-scan dual-Doppler coverage between RaXPol and the MWR-05XP enabled some storm-scale features to be resolved during the intensification of the second tornado, including the wind-field around a weak reflectivity band associated with a horizontal vortex that was viewed by the RaXPol crew.