Supporting data for "Evaluating Tools for Diagnosis and Nowcasting Precipitation Type and Freezing Rain: Results from the 3–4 February 2022 Winter Storm in the Hudson Valley"

User Guide Supporting data for "Supporting data for "Evaluating Tools for Diagnosis and Nowcasting Precipitation Type and Freezing Rain: Results from the 3–4 February 2022 Winter Storm in the Hudson Valley" Version 2.0 21 November 2025 This dataset contains supporting data associated with the publication Evaluating Tools for Diagnosis and Nowcasting Precipitation Type and Freezing Rain: Results from the 3–4 February 2022 Winter Storm in the Hudson Valley (Minder et al. 2025). Specifically, it includes data that are not otherwise available at the time of publication. For access to other datasets used in the publication, please consult its Data Availability Statement. Below is basic information about each dataset and citations 1. Random forest (RF) machine learning model p-type time series This data is from the random forest (RF) machine learning model described in Filipiak et al. (2023). This model produces probabilistic analyses and forecasts of winter p-type, trained on manual citizen science p-type observations from the Community Collaborative Rain, Hail and Snow network. The model ingests NYSM station observations, operational upper-air sounding observations, and short-term vertical profile forecasts from the operational North American Model nest (NAMNEST, 3km horizontal grid) as input features. From these features, the RF predicts probability of RA, SN, FZRA, and PL. In this study, we focus on analysis of the dominant (most probable) p-type predicted by the RF model. • Files: o RF_PROF_ALBA.csv: RF model time series for location of NYSM PROF_ALBA station. o RF_PROF_REDH.csv: RF model time series for location of NYSM PROF_REDH station. • Format: Comma separated values, with header giving variable names and units. Missing data is indicated with values of NaN. The columns provide include: o Latitude and longitude (in degrees) o Probabilities of each diagnosed p-type (in percentages) § The reporting notation diWers slightly from what is used in the paper. In particular: SL (sleet) = PL (ice pellets). o The dominant (most probable) p-type o The time valid time (in UTC) • Citation: o Filipiak, B. C., N. P. Bassill, K. L. Corbosiero, A. L. Lang, and R. A. Lazear, 2023: Probabilistic Forecasting Methods of Winter Mixed-Precipitation Events in New York State Utilizing a Random Forest. Artif. Intell. Earth Syst., 2, e220080, https://doi.org/10.1175/AIES-D-22-0080.1. 2. New York State Mesonet microwave radiometer (MWR) p-type time series This data is from the New York State Mesonet (NYSM) microwave radiometer (MWR) p-type diagnostic, using a modified parcel thickness method, as described in Shrestha et al. (2023). • Files: • MWR_PROF_ALBA_v2.csv: MWR p-type diagnosis based on data from the NYSM PROF_ALBA station. • MWR_PROF_REDH_v2.csv: MWR p-type diagnosis based on data from the NYSM PROF_REDH station. • Format: Coma separated values, with header giving variable names and units. The columns provide include: o Timestamp (in UTC) o ptype: all p-types diagnosed at the given time, potentially including mixtures (e.g., “FZR/SLT”) • Notes: • In p-type reports, the abbreviations diWer from what is used in the paper. In particular: FZR = FZRA, SLT= PL. • Citation: • Shrestha, B., J. Wang, J. A. Brotzge, and N. Bain, 2023: Winter Precipitation Type from Microwave Radiometers in New York State Mesonet Profiler Network. Wea. Forecasting, 38, 1563–1574, https://doi.org/10.1175/WAF-D-23-0035.1. 3. Spectral Bin Classifier (SBC) p-type and Freezing Rain Accumulation National Analysis (FRANA) This data is from the recently developed spectral bin classifier (SBC) and Freezing Rain Accumulation National Analysis (FRANA). The SBC is a tool for gridded diagnoses of precipitation phase, described in Reeves et al. (2016). It is a one-dimensional binmicrophysics model that uses thermodynamic profiles to explicitly compute the liquid-water fraction (LWF) of individual hydrometeors as they descend from the top of the cloud to the ground. The SBC relies on inputs of the vertical profiles of temperature, humidity, and pressure from an NWP model. The SBC analyses used here were run as part of an experimental version of the Multi-Radar/Multi-Sensor (MRMS) system using 1-hr HRRR forecasts. FRANA (Tripp et al. 2025) relies on SBC to denote areas of FZRA and the MRMS quantitative precipitation estimates for precipitation rate. It uses the Freezing Rain Accumulation Model (FRAM; Sanders and Barjenbruch 2016) equations to calculate ice-to-liquid ratios based on HRRR-analyses of 10-m winds and wet bulb temperature. • Files: • FRANA_grids_20220205_0000_48h.nc: Gridded FRANA analysis of 48-h freezing rain accumulation, valid for 0000 UTC 03 February 2022 – 0000 UTC 05 February 2022. • sbc_grids_[yyyymmdd_HHMM].nc: Gridded SBC analysis of p-type, valid for the 1-hour period ending at the noted valid time. • SBC_FRANA_time_series.csv: Hourly time series of SBC p-type and FRANA freezing rain accumulation at locations of KALB ASOS station and REDH NYSM station. • Format: • FRANA_grids_20220205_0000_48h.nc: netCDF file with embedded variable metatdata • sbc_grids_[yyyymmdd_HHMM].nc: netCDF file with embedded variable metatdata • SBC_FRANA_time_series.csv: Comma separated values, with header giving variable names and units. • Citations: • Reeves, H. D., A. V. Ryzhkov, and J. Krause, 2016: Discrimination between Winter Precipitation Types Based on Spectral-Bin Microphysical Modeling. J. Appl. Meteor. Clamato., 55, 1747–1761, https://doi.org/10.1175/JAMC-D-16-0044.1. • Sanders, K. J., and B. L. Barjenbruch, 2016: Analysis of Ice-to-Liquid Ratios during Freezing Rain and the Development of an Ice Accumulation Model. Wea. Forecasting, 31, 1041–1060, https://doi.org/10.1175/WAF-D15-0118.1. • Tripp, D. D., A. D. Werkema, H. D. Reeves, B. L. Barjenbruch, and K. J. Sanders, 2025: Creation and Evaluation of the Freezing Rain Accumulation National Analysis (FRANA) in Preparation for NWS Operations. Wea. Forecasting, 40, 319–332, https://doi.org/10.1175/WAF-D-24-0085.1. 4. Version history • V1: original • V2: Updated MWR profiler-based p-type (MWR_PROF_ALBA_v2.csv, MWR_PROF_REDH_v2.csv) to use latest methods and to remove "greatest severity bias" merging of p-types.