Impact of Assimilating New York Stat MesoNet and Two Offshore LiDARs on Modeling the New York Bight Jet

The New York Bight jet is a southerly low-level (<200 m above the surface) wind maximum (>11 m/s) that primarily develops during the warm season off the northern New Jersey coast and south of Long Island. Offshore wind power would benefit from these enhanced winds, though it requires accurate predictions of low-level jet (LLJ) events in the lowest few hundred meters. There are new observations to help with the prediction of these jet events. Two Light Detection and Ranging (LiDAR) buoys were deployed by the New York State Energy Research and Development Authority (NYSERDA) a few years ago within the NY Bight region to provide high fidelity wind measurements. There is also the New York State Mesonet (NYSM) with 126 surface stations and 17 profiler sites (including scanning LiDARs) around the state.

This presentation highlights the impact of assimilating the NYSM and NYSERDA offshore LiDAR data for a short-range (12 h) ensemble forecast of a LLJ event on 28-29 June 2021, which had observed windspeed maximum of ~17 m s^-1 at ~160 m ASL around 100 km offshore. The Data Assimilation Research Testbed (DART; Anderson et al. 2009) is used in conjunction with a 40-member Weather Research and Forecasting (WRF; Skamarock et al. 2008) ensemble at 9-km grid spacing to assimilate observations every six hours starting with the Global Forecast System (GFS) analysis on 0600 UTC 25 June 2021. The first experiment (called “Conventional-only”) only assimilated conventional observations (radiosondes, aerodrome reports, ships and buoys, aircraft, and NOAA profilers). Two additional experiments were run in which more observations were assimilated, one which included the NY Mesonet observations in addition to the conventional observations (“Conventional-Mesonet”), another which then also included the two offshore LiDARs (“Conventional-Mesonet-LiDARs"). In each of these experiments, ten random members from the 1800 UTC 28 June 2021 cycle are used to initialize WRF free runs at 9 km grid spacing. Both the data assimilation cycles and free runs use model physics similar to the RAP model, including the local MYNN 2.5 planetary boundary layer (PBL) scheme. For comparison, additional free runs are initialized with the RAP and GFS analyses starting at 1800 UTC 28 June 2021.

During the first 2-3 forecast hours, the free run from the Conventional-Mesonet-LiDARs experiment has the smallest wind speed error of the LLJ maximum (~-0.8 m s^-1 or -6% of the observation). However, by 6-12 h into the forecast, the ensemble-mean of that run underpredicts the intensifying LLJ by 2-4 m·s^-1 (12-23%). The free runs from the Conventional-only and Conventional-Mesonet experiments have comparable wind speed errors after 6 h. Meanwhile, the GFS-initialized free run initially underpredicts the wind speeds by 1.5-2.0 m·s^-1 (11-15%), but then intensifies the LLJ such that the errors after 6 h are 0.5-1 m·s^-1 (25-30%) smaller than the Convection-Mesonet-LiDAR free run. The RAP free run overall has the smallest wind speed errors after 9 h (~1 m·s^-1 or ~30% smaller than the GFS). Alternate non-local PBL schemes, such as the YSU, are also tested and excessively mix-out the wind profile, resulting in larger wind errors in all runs. Overall, this case shows large sensitivity to both the initial conditions and the PBL parameterization for these jet events. Reasons for the differences in the runs will be highlighted.

References:

Anderson, J., T. Hoar, K. Raeder, H. Liu, N. Collins, R. Torn, A. Avellano, A., 2009: The data assimilation research testbed: A community facility. Bull. Am. Meteorol. Soc., 90, 1283–1296. doi:10.1175/2009BAMS2618.1.

Skamarock, W., J. B. Klemp, J. Dudhia, D. O. Gill, D. Barker, M. G. Duda, X. -Y. Huang, and W. Wang, 2008: A Description of the Advanced Research WRF Version 3. NCAR Technical Note NCAR/TN-475+STR. doi:10.5065/D68S4MVH.