Subtropical South American Hailstorm Characteristics and Environments.

Hailstorms in subtropical South America are known to be some of the most frequent anywhere in the world, causing significant damage to the local agricultural economy every year. Convection in this region tends to be orographically forced, with moisture supplied from the Amazon rainforest by the South American low-level jet. Previous climatologies of hailstorms in this region have been limited to localized and sparse observational networks. Due to the lack of sufficient ground-based radar coverage, objective radar-derived hail climatologies have also not been produced for this region. As a result, this study uses a 16-year dataset of TRMM Precipitation Radar and Microwave Imager observations to identify possible hailstorms remotely, using 37-GHz brightness temperature as a hail proxy. By combining satellite instruments and ERA-Interim reanalysis data, this study produces the first objective study of hailstorms in this region. Hailstorms in subtropical South America have an extended diurnal cycle, often occurring in the overnight hours. Additionally, they tend to be multi-cellular in nature, rather than discrete. High-probability hailstorms (≥ 50% probability of containing hail) tend to be deeper by 1-2 km and horizontally larger by greater than 15,000 km2 than storms having a low-probability of containing hail (< 25% probability of containing hail). Finally, hailstorms are supported synoptically by strong upper- and lower-level jets, anomalously warm and moist low levels, and enhanced instability. The findings of this study will support the forecasting of these severe storms and mitigation of their damages within this region.

Tropical Cyclone Intensity Estimation Using a Deep Convolutional Neural Network.

Tropical cyclone intensity estimation is a challenging task as it required domain knowledge while extracting features, significant pre-processing, various sets of parameters obtained from satellites, and human intervention for analysis. The inconsistency of results, significant pre-processing of data, complexity of the problem domain, and problems on generalizability are some of the issues related to intensity estimation. In this study, we design a deep convolutional neural network architecture for categorizing hurricanes based on intensity using graphics processing unit. Our model has achieved better accuracy and lower root-mean-square error by just using satellite images than 'state-of-the-art' techniques. Visualizations of learned features at various layers and their deconvolutions are also presented for understanding the learning process.

Hurricane Imaging Radiometer (HIRAD) Wind Speed Retrievals and Validation Using Dropsondes.

Surface wind speed retrievals have been generated and evaluated using Hurricane Imaging Radiometer (HIRAD) measurements from flights over Hurricane Joaquin, Hurricane Patricia, Hurricane Marty, and the remnants of Tropical Storm Erika, all in 2015. Procedures are described here for producing maps of brightness temperature, which are subsequently used for retrievals of surface wind speed and rain rate across a ~50 km wide swath for each flight leg. An iterative retrieval approach has been developed to take advantage of HIRAD's measurement characteristics. Validation of the wind speed retrievals has been conducted, using 636 dropsondes released from the same WB-57 high altitude aircraft carrying HIRAD during the Tropical Cyclone Intensity (TCI) experiment. The HIRAD wind speed retrievals exhibit very small bias relative to the dropsondes, for winds tropical storm strength (17.5 m s-1) or greater. HIRAD has reduced sensitivity to winds weaker than tropical storm strength, and a small positive bias (~2 m s-1) there. Two flights with predominantly weak winds according to the dropsondes have abnormally large errors from HIRAD, and large positive biases. From the other flights, root mean square differences between HIRAD and the dropsonde winds are 4.1 m s-1 (33%) for winds below tropical storm strength, 5.6 m s-1 (25%) for tropical storm strength winds, and 6.3 m s-1 (16%) for hurricane strength winds. Mean absolute differences for those categories are 3.2 m s-1 (25%), 4.3 m s-1 (19%), and 4.8 m s-1 (12%), with bias near zero for tropical storm and hurricane strength winds.

Automated Storm Tracking and the Lightning Jump Algorithm Using GOES-R Geostationary Lightning Mapper (GLM) Proxy Data.

This study develops a fully automated lightning jump system encompassing objective storm tracking, Geostationary Lightning Mapper proxy data, and the lightning jump algorithm (LJA), which are important elements in the transition of the LJA concept from a research to an operational based algorithm. Storm cluster tracking is based on a product created from the combination of a radar parameter (vertically integrated liquid, VIL), and lightning information (flash rate density). Evaluations showed that the spatial scale of tracked features or storm clusters had a large impact on the lightning jump system performance, where increasing spatial scale size resulted in decreased dynamic range of the system's performance. This framework will also serve as a means to refine the LJA itself to enhance its operational applicability. Parameters within the system are isolated and the system's performance is evaluated with adjustments to parameter sensitivity. The system's performance is evaluated using the probability of detection (POD) and false alarm ratio (FAR) statistics. Of the algorithm parameters tested, sigma-level (metric of lightning jump strength) and flash rate threshold influenced the system's performance the most. Finally, verification methodologies are investigated. It is discovered that minor changes in verification methodology can dramatically impact the evaluation of the lightning jump system.

Tropical Cyclone Diurnal Cycle as Observed by TRMM.

Previous work has indicated a clear, consistent diurnal cycle in rainfall and cold cloudiness coverage around tropical cyclones. This cycle may have important implications for structure and intensity changes of these storms and the forecasting of such changes. The goal of this paper is to use passive and active microwave measurements from the Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI) and Precipitation Radar (PR), respectively, to better understand the tropical cyclone diurnal cycle throughout a deep layer of a tropical cyclone's clouds. The composite coverage by PR reflectivity ≥20 dBZ at various heights as a function of local standard time (LST) and radius suggests the presence of a diurnal signal for radii <500 km through a deep layer (2-10 km height) of the troposphere using 1998-2011 Atlantic tropical cyclones of at least tropical storm strength. The area covered by reflectivity ≥20 dBZ at radii 100-500 km peaks in the morning (0130-1030 LST) and reaches a minimum 1030-1930 LST. Radii between 300-500 km tend to reach a minimum in coverage closer to 1200 LST before reaching another peak at 2100 LST. The inner core (0-100 km) appears to be associated with a single-peaked diurnal cycle only at upper levels (8-10 km) with a maximum at 2230-0430 LST. The TMI rainfall composites suggest a clear diurnal cycle at all radii between 200 and 1000 km with peak rainfall coverage and rain rate occurring in the morning (0130-0730 LST).

Properties of hail storms over China and the United States from the Tropical Rainfall Measuring Mission.

A 16-yr record of hail reports over the south U.S. and from weather stations in China are collocated with Precipitation Features (PF) derived from the Tropical Rainfall Measuring Mission (TRMM) radar and passive microwave observations. Differences in the way hail is reported in the two nations make it difficult to draw meaningful conclusions about storm frequency. But taking the two together yields a wide spectrum of hail sizes, suitable for comparing with remote sensing measurements. While U.S. hail reports are dominated by cases with hail size greater than 19 mm, hail reports in China mostly include diameters of 1-10 mm and mostly occur over the Tibetan Plateau. The fraction of PFs collocated with hail reports (hail PFs) reaches 3% in the plains of the U.S. In China, the fraction is higher in high elevation regions than low elevation regions. Hail PFs (as reported in the U.S.) show lower brightness temperatures, higher lightning flash rates, stronger maximum reflectivity, and higher echo tops than those with smaller hail, as reported in China. The average near surface maximum reflectivity of hail PFs at high elevations (≥ 2000 m) in China is about 5 dB smaller than those at low elevations. Larger hail is reported with PFs having stronger maximum reflectivity above 6 km, though the median of maximum reflectivity values at levels below 5 km is similar among the storms with large and small hail sizes.