The Current Icing Product (CIP; Bernstein et al. 2005) and Forecast Icing Product (FIP; Wolff et al. 2009) were originally developed by the United States’ National Center for Atmospheric Research (NCAR) under sponsorship of the Federal Aviation Administration (FAA) in the mid 2000’s and provide operational icing guidance to users through the NOAA Aviation Weather Center (AWC). The current operational version of FIP uses the Rapid Refresh (RAP; Benjamin et al. 2016) numerical weather prediction (NWP) model to provide hourly forecasts of Icing Probability, Icing Severity, and Supercooled Large Drop (SLD) Potential. Forecasts are provided out to 18 hours over the Contiguous United States (CONUS) at 15 flight levels between 1,000 ft and FL290, inclusive, and at a 13-km horizontal resolution. CIP provides similar hourly output on the same grid, but utilizes geostationary satellite data, ground-based radar data, Meteorological Terminal Air Reports (METARS), lightning data, and voice pilot reports (PIREPs) in addition to the RAP model output to provide near-realtime icing guidance. This paper presents recent enhancements to the prototype versions of CIP and FIP (CIP v2.0 and FIP v2.0, respectively). The enhancements described are intended to take better advantage of enhanced model resolution and microphysics parameterization as well as state-of-the-art observations for icing diagnosis and forecasting.

Abstract. The lack of shortwave (SW, visible, and near-infrared) geostationary satellite data at night results in degradation of many weather forecasts and real-time diagnostic products. We present a method to extrapolate SW GOES-16 advanced baseline imager data through night using nighttime longwave (LW, infrared) observations and the relationships between LW and SW data observed during the previous day. The method is not a forecast since it requires LW nighttime observations but can provide continuity through day, night, and satellite terminator hours. To provide performance statistics, the algorithm is applied during the day so the SW extrapolations can be compared to observations. Typical mean absolute errors (MAEs) range from 1.0% to 12.7% reflectance depending on the SW channel. These MAEs can be predicted using a diagnostic metric called 0-h MAE which quantifies the quality of the algorithm’s input data. In addition to quantitative error statistics, three case studies are presented, including an animation of extrapolated imagery from dusk through dawn. Considerations for future improvements include use of convolutional neural networks and/or object-based extrapolations where mesoscale features are extrapolated individually.

The National Center for Atmospheric Research (NCAR) recently updated the comprehensive wind power forecasting system in collaboration with Xcel Energy addressing users’ needs and requirements by enhancing and expanding integration between numerical weather prediction and machine-learning methods. While the original system was designed with the primary focus on day-ahead power prediction in support of power trading, the enhanced system provides short-term forecasting for unit commitment and economic dispatch, uncertainty quantification in wind speed prediction with probabilistic forecasting, and prediction of extreme events such as icing. Furthermore, the empirical power conversion machine-learning algorithms now use a quantile approach to data quality control that has improved the accuracy of the methods. Forecast uncertainty is quantified using an analog ensemble approach. Two methods of providing short-range ramp forecasts are blended: the variational doppler radar analysis system and an observation-based expert system. Extreme events, specifically changes in wind power due to high winds and icing, are now forecasted by combining numerical weather prediction and a fuzzy logic artificial intelligence system. These systems and their recent advances are described and assessed.

The High-Resolution Rapid Refresh (HRRR) model with its hourly updating cycles provides multiple weather forecasts valid at any given time. A logical combination of these individual deterministic forecasts is postulated to show more skill than any single forecast for predicting clouds containing supercooled liquid water (SLW), an aircraft icing threat. To examine the potential value of using multiple HRRR forecasts for icing prediction, a time-lag-ensemble (TLE) averaging method of combining a number of HRRR forecasts was implemented for a multiple month real-time test during the winter of 2016/17. The skills of individual HRRR and HRRR-TLE aircraft icing predictions were evaluated using icing pilot reports (PIREPs) and surface weather observations and compared with the operational Forecast Icing Product (FIP) using the Rapid Refresh (RAP) model. The HRRR-TLE was found to produce a higher capture rate of icing PIREPs and surface icing conditions of freezing drizzle or freezing rain than single deterministic HRRR forecasts. As a trade-off, the volume of airspace warned in HRRR-TLE increased, resulting in a higher false detection rate than in the deterministic HRRR forecasts. Overall, the HRRR-TLE had similar probability of detection and volume of airspace warned for icing as the operational FIP prediction for the icing probability of 25% or greater. Alternative techniques for composing TLE from multiple HRRR forecasts were tested in postseason rerun experiments. The rerun tests also included a comparison of the skills of HRRR and HRRR-TLE to the skills of RAP and RAP-TLE.

The accurate diagnosis of in-flight icing conditions is dependent on surface observations of cloud coverage, cloud base height, and surface precipitation type. However, the network for collecting these data over the the United States is neither contiguous nor evenly distributed. Surface observational gaps exist over much of the domain where in-flight icing conditions are diagnosed. Due to the way these observations are treated when diagnosing inflight icing conditions, the result is often circles where icing conditions are possible next to areas with no icing conditions diagnosed due to absence of surface observations. To avoid these visually unappealing and scientifically inconsistent artifacts, a method was developed to create surrogate surface observation data from numerical weather prediction model output. Using the individual condensate fields, accumulated precipitation, and temperature all three of the required datasets used from surface observations in diagnosing in-flight icing conditions were derived. The Current Icing Product (CIP) shows in-flight icing diagnoses created with the model derived surface observations that are similar to those created when using only real observations.

In-flight icing is a significant hazard in Alaska as the atmospheric environment is complex and ranges from maritime to continental and temperate to polar. An analysis of radiosonde data conditions for different climate zones reveals a high frequency of icing conditions year-round, varying with season and altitude. Many locations in Alaska depend on air travel for transportation, especially in smaller aircraft that fly at icing-prone altitudes. Thus, accurate diagnoses and forecasts of the icing environment, tuned to these varying conditions, are needed. Icing products are currently under development that are anticipated to meet the needs of aviation users in Alaska. The forecast product will be available first and is based on the Forecast Icing Product, originally developed for use in the CONUS, and predicts icing probability, supercooled large drop potential, and severity. The current spatial resolution is 13 km; high-resolution (3-km) model runs have also been used in the Alaska forecast algorithm to assess their value. An icing diagnosis algorithm that combines observations with model output, much like the Current Icing Product, is also in early development. To improve that product, the use of polar orbiting satellite data is being explored. These observations may be added to the diagnosis algorithm to provide observations where geostationary satellite data are not available.

Convectively induced turbulence (CIT) has been shown to cause or factor into a large portion of weather-related commercial aviation accidents. Determining areas of CIT is difficult since CIT is a relatively small scale phenomenon. The Federal Aviation Administration issued guidelines for pilots to avoid thunderstorms, but flying around a storm can waste time and money. In-cloud CIT is created by dynamics within the cloud, such as the updraft. These same dynamics promote cloud electrification and subsequently, the generation of lightning. Therefore, lightning may be an indicator of a robust updraft and the likelihood of CIT. With the expected increase in availability of global lightning data through the launch of the GOES-R satellite, this relationship could improve the identification of CIT in otherwise data-sparse locations. Data from the NCAR Turbulence Detection Algorithm were compared with total lightning data measured by the Colorado Lightning Mapping Array and dual-polarimetric radar data from the Denver, Colorado and Cheyenne, Wyoming WSR-88Ds. This was done in order to determine possible temporal and spatial relationships of turbulence to electrical and microphysical storm properties. In several case studies of severe storms over Colorado, Wyoming, and Nebraska, it was observed that higher total lightning frequencies accompanied higher turbulence intensities. CIT often occurred prior to any lightning discharges. Likewise, lightning was located within the mixed phase region of a storm, while turbulence maxima often extended just above this region. Additionally, turbulence was observed within the storm after the last lightning strike. Results suggest lightning may be indicative of in-cloud CIT.

The goal of this study is to develop an Icing Hazard Level Algorithm (IHLA) that utilizes dual-polarization radar data and output from an operational numerical weather prediction model. This work is motivated by the upcoming dual-polarization upgrade to the NEXRAD network. The IHLA algorithm is a modular design consisting of melting level detection, temperature profile adjustment and currently 3 icing condition detection modules.

The goal of icing forecasting research being conducted at the National Center for Atmospheric Research (NCAR) for the Federal Aviation Administration’s (FAA) Aviation Weather Research Program (AWRP) is to provide timely and accurate forecasts of inflight icing conditions. The flying public wants to know not only where icing conditions are likely to reside, but also the probability of their occurrence and expected severity. Automated diagnosis and forecast icing products have been developed at NCAR and deployed at the Aviation Weather Center (AWC), where they provide this information to pilots, forecasters, and dispatchers. These products, known as the Current and Forecast Icing Products (CIP and FIP, respectively), have been approved for operational decision making by these groups. Recently, changes were made to both algorithms to accommodate the transition in numerical weather prediction (NWP) models, from the Rapid Update Cycle (RUC) to the Weather Research and Forecasting Rapid Refresh (WRF-RAP). This transition required some changes to the algorithms to handle updated model information and a verification of the results. The verification study confirmed that the new model had the desired effects on the icing products and also brought to light some interesting information on the handling of convection and supercooled large drops (SLD). During this transition other upgrades and changes were also made to the algorithms dealing with icing severity at night, the use of radar data, and the development of an algorithm testbed.

The Current Icing Product (CIP) uses five observational datasets in addition to numerical weather prediction (NWP) model data to diagnose a three-dimensional icing probability and severity. By default, there are thresholds that limit the age of the observational data that CIP uses. If the data time does not meet the threshold and it is a required dataset (surface observations, satellite), then CIP will not run. By utilizing the probability of detection (POD) value of only the positive icing pilot reports (PODy) for both of the required input datasets, it was found that the current thresholds are adequate. Satellite data that are 30 minutes old produce a PODy of 0.82, but that drops to 0.78 when the age of the satellite data reaches 120 minutes. Similar results were found using surface observation data ranging from 60 to 180 minutes old. Understanding the behavior and performance of CIP when it is required to use less than optimal input data is crucial for making improvements to the algorithm and testing additional datasets. A methodology is presented to assist with these tasks in the future, as well as a more in depth look at the effect of varying input dataset age has on PODy of both CIP icing probability and severity.

As an aircraft flies through supercooled liquid water, the liquid freezes instantaneously to the airframe thus altering its lift, drag, and weight characteristics. In-flight icing is a contributing factor to many aviation accidents, and the reliable detection of this hazard is a fundamental concern to aviation safety. The scientific community has recently developed products to provide in-flight icing warnings. NASA's Icing Remote Sensing System (NIRSS) deploys a vertically--pointing Ka--band radar, a laser ceilometer, and a profiling multi-channel microwave radiometer for the diagnosis of terminal area in-flight icing hazards with high spatial and temporal resolution. NCAR s Current Icing Product (CIP) combines several meteorological inputs to produce a gridded, three-dimensional depiction of icing severity on an hourly basis. Pilot reports are the best and only source of information on in-situ icing conditions encountered by an aircraft. The goal of this analysis was to ascertain how the testbed NIRSS icing severity product and the operational CIP severity product compare to pilot reports of icing severity, and how NIRSS and CIP compare to each other. This study revealed that the icing severity product from the ground-based NASA testbed system compared very favorably with the operational model-based product and pilot reported in-situ icing.