Identifying Meteorological Drivers for Errors in Modeled Winds along the Northern California Coast

Ye Liu aPacific Northwest National Laboratory, Richland, Washington

Search for other papers by Ye Liu in
Current site
Google Scholar
PubMed
Close
https://orcid.org/0000-0001-5131-8412
,
Brian Gaudet aPacific Northwest National Laboratory, Richland, Washington

Search for other papers by Brian Gaudet in
Current site
Google Scholar
PubMed
Close
,
Raghavendra Krishnamurthy aPacific Northwest National Laboratory, Richland, Washington

Search for other papers by Raghavendra Krishnamurthy in
Current site
Google Scholar
PubMed
Close
,
Sheng-Lun Tai aPacific Northwest National Laboratory, Richland, Washington

Search for other papers by Sheng-Lun Tai in
Current site
Google Scholar
PubMed
Close
,
Larry K. Berg aPacific Northwest National Laboratory, Richland, Washington

Search for other papers by Larry K. Berg in
Current site
Google Scholar
PubMed
Close
,
Nicola Bodini bNational Renewable Energy Laboratory, Golden, Colorado

Search for other papers by Nicola Bodini in
Current site
Google Scholar
PubMed
Close
,
Alex Rybchuk bNational Renewable Energy Laboratory, Golden, Colorado

Search for other papers by Alex Rybchuk in
Current site
Google Scholar
PubMed
Close
, and
Andrew Kumler bNational Renewable Energy Laboratory, Golden, Colorado

Search for other papers by Andrew Kumler in
Current site
Google Scholar
PubMed
Close
Free access

Abstract

An accurate wind resource dataset is required for assessing the potential energy yield of floating offshore wind farms that are expected along the California outer continental shelf. The National Renewable Energy Laboratory has developed and disseminated an updated wind resource dataset offshore of California, using the Weather Research and Forecasting Model, referred to as the CA20 dataset. As compared to buoy lidar measurements that have become available recently, the CA20 dataset showed significant positive biases for 100-m wind speeds along Northern California wind energy lease areas. To investigate the meteorological drivers for the model errors, we first consider two 1-yr simulations run with two different planetary boundary layer (PBL) parameterizations: the Mellor–Yamada–Nakanishi–Niino (MYNN) PBL scheme (the chosen configuration in the CA20 dataset) and the Yonsei University PBL scheme (which significantly reduces the bias in modeled winds). By comparing the 1-yr simulations to the concurrent lidar buoy observations, we find that errors are larger with the MYNN PBL scheme in warm seasons. We then dive deeper into the analysis by running simulations for short-term (3-day) case studies to evaluate the sensitivity of initial/boundary condition forcings on model results. By analyzing the short-term simulations, we find that during synoptic-scale northerly flows driven by the North Pacific high and inland thermal low, a coastal warm bias in the MYNN simulation is mainly responsible for the modeled wind speed bias by altering the boundary layer thermodynamics. The results of our analysis will help guide the creation of an updated version of the CA20 dataset.

© 2024 American Meteorological Society. This published article is licensed under the terms of the default AMS reuse license. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Ye Liu, ye.liu@pnnl.gov

Abstract

An accurate wind resource dataset is required for assessing the potential energy yield of floating offshore wind farms that are expected along the California outer continental shelf. The National Renewable Energy Laboratory has developed and disseminated an updated wind resource dataset offshore of California, using the Weather Research and Forecasting Model, referred to as the CA20 dataset. As compared to buoy lidar measurements that have become available recently, the CA20 dataset showed significant positive biases for 100-m wind speeds along Northern California wind energy lease areas. To investigate the meteorological drivers for the model errors, we first consider two 1-yr simulations run with two different planetary boundary layer (PBL) parameterizations: the Mellor–Yamada–Nakanishi–Niino (MYNN) PBL scheme (the chosen configuration in the CA20 dataset) and the Yonsei University PBL scheme (which significantly reduces the bias in modeled winds). By comparing the 1-yr simulations to the concurrent lidar buoy observations, we find that errors are larger with the MYNN PBL scheme in warm seasons. We then dive deeper into the analysis by running simulations for short-term (3-day) case studies to evaluate the sensitivity of initial/boundary condition forcings on model results. By analyzing the short-term simulations, we find that during synoptic-scale northerly flows driven by the North Pacific high and inland thermal low, a coastal warm bias in the MYNN simulation is mainly responsible for the modeled wind speed bias by altering the boundary layer thermodynamics. The results of our analysis will help guide the creation of an updated version of the CA20 dataset.

© 2024 American Meteorological Society. This published article is licensed under the terms of the default AMS reuse license. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Ye Liu, ye.liu@pnnl.gov

1. Introduction

In 2022, the U.S. offshore wind energy project development and operational pipeline grew to a potential generating capacity of 40 000 MW, with the significant contribution of capacity in the outer continental shelf off the coast of California (Musial et al. 2022). The deep waters offshore of California will require wind turbines installed on floating platforms anchored to the seabed. To provide accurate cost estimates for floating wind energy, the National Renewable Energy Laboratory released a new 20-yr (2000–19) wind resource dataset (referred to as CA20) that replaces the Wind Integration National Dataset (WIND) Toolkit (Draxl et al. 2015) for use in the California outer continental shelf (Optis et al. 2020). The CA20 simulation configuration improved upon the WIND Toolkit setup by spanning a longer period and using more modern version of model physics and forcing data, and it was validated against measurements available at the time, namely, coastal radars and near-surface buoy measurements (Olson et al. 2019a; Optis et al. 2020; Hersbach et al. 2020). However, the comparison with two recently available floating lidars revealed large positive wind speed biases at all considered heights (Bodini et al. 2022). The errors are significantly greater than comparable datasets have found in other regions (Hahmann et al. 2020) and could significantly hinder application of the data for subsequent scientific analyses to evaluate the offshore wind resource. This study will address two questions: “What factors contribute to the modeled wind speed errors within the boundary layer?” and “How are the errors associated with the atmospheric conditions?”

Among many other processes and factors, the planetary boundary layer (PBL) scheme, which controls how turbulence distributes momentum in the atmosphere and strongly influences the wind shear profiles, is a critical parameterization in numerical weather forecasting models. Prior to choosing the model configuration for the CA20 dataset, a thorough sensitivity study was performed by varying PBL schemes, reanalysis forcing, sea surface temperature forcing, and surface layer schemes (Optis et al. 2020). The model showed the largest sensitivity when varying the PBL schemes. The two PBL schemes evaluated were the Mellor–Yamada–Nakanishi–Niino (MYNN; Nakanishi and Niino 2009) scheme and the Yonsei University (YSU; Hong et al. 2006) scheme (Optis et al. 2020). The MYNN scheme produced a moderately smaller bias (0.28 m s−1 for MYNN and 0.36 m s−1 for YSU) and a better probability density function than the YSU scheme when compared to surface buoy [∼4 m above mean sea level (MSL)] and coastal radar wind profiler observations. Similar model biases were also observed along the Atlantic coast (Pronk et al. 2022). Therefore, the MYNN PBL scheme was chosen as a default configuration for the CA20 dataset. Note that, the MYNN PBL scheme is widely used in various wind atlases and weather forecasting models, and significant improvements have been made to the scheme (Liu et al. 2022; Berg et al. 2019, 2021; Hahmann et al. 2020; Olson et al. 2019b), while YSU scheme was used in the WIND Toolkit (Draxl et al. 2015).

One main difference between the MYNN and YSU schemes is that the former is a local closure PBL scheme whereas the latter is a nonlocal scheme. In the upper part of the convective PBL, the heat flux determined by large eddies is often counter to the local gradient, with upward heat flux arising from the nonlocal transport by buoyant plumes that initiated near the surface (Deardorff 1972). An eddy-diffusion mass-flux convective approach was included in MYNN to better account for the nonlocal thermals (Soares et al. 2004; Olson et al. 2019b), which significantly improved the forecast of wind-turbine-height wind speed compared to the default MYNN scheme in one study (e.g., Olson et al. 2019a). In this study, we evaluate the performance of the two PBL schemes in modeling hub-height wind speeds, and one step further, investigate the physical mechanisms related to hub-height winds.

The regional climate and mountainous terrain along the coast of California increase the complexity of modeling turbine hub-height wind speed and could potentially contribute to the CA20 biases. Typical spring and summer conditions offshore of California consist of strong northerly low-level winds, which are driven by factors including the North Pacific high (NPH), low pressure over the Southwest United States, well-mixed marine boundary layer (MBL), oceanic upwelling, and a meridional coastline with coastal mountains, capes, and hydraulic MBL dynamics (Rahn et al. 2014; Beardsley et al. 1987; Dorman and Koračin 2008; Fewings et al. 2016). The strong land–sea thermal contrast further accelerates the coastal northerly winds, establishing a low-level jet along the coast (e.g., Burk and Thompson 1996). A maximum wind speed around and offshore of coastal capes is often observed or modeled (e.g., Dorman and Koračin 2008; Smith et al. 2018; Fewings et al. 2016; Parish et al. 2016). Prevailing northerly winds are frequently interrupted by westerlies associated with atmospheric rivers (Guan and Waliser 2015; Dettinger et al. 2011; Payne et al. 2020), cold fronts, or winter storms. The objective of this study is to investigate the physical drivers of the CA20 model biases with a focus on analyzing physical processes connected to the interaction between the offshore environment and the lower atmospheric boundary layer. To achieve this goal, a pair of 1-yr simulations using the CA20 model configuration with YSU and MYNN PBL schemes are conducted along with five short-term (3-day) simulations (with varying configurations) to assess the model sensitivities to model setup that influence hub-height winds biases. The analyses from these simulations reveal the seasonal variation of the wind speed errors and some physical mechanisms that are connected to the hub-height wind speed biases.

2. Data and method

a. Lidar buoy and land-based radar data

The U.S. Department of Energy in collaboration with the Bureau of Ocean Energy Management deployed two buoys equipped with Doppler lidars along the coast of California to directly observe the offshore wind resource. The buoys were deployed off the coasts of Morro Bay and Humboldt within the wind energy lease areas (Krishnamurthy et al. 2023). These buoys were deployed for an entire annual cycle and provided measurements of wind and turbulence (from the surface to 250 m MSL), surface meteorology, sea surface temperature, solar radiation, two-dimensional wave spectra, and ocean current profile of speed and direction (Krishnamurthy and Sheridan 2020a,b, 2021a,b). The Humboldt lidar had to be serviced due to a failure in the buoy power system and its observations are available from October 2020 to December 2020 and from June 2021 to December 2021, whereas the Morro Bay buoys were in service throughout the whole period. The accuracy of the lidar measurements can be potentially affected by precipitation events and foggy conditions (Newsom and Krishnamurthy 2022), although the uncertainty of lidar measurements during such conditions is a current topic of research. After carefully evaluating the time-height plots of carrier to noise ratio and horizontal wind speed during the analyzed periods, we find the time series does not show any systematic issues with the observations.

We also collect concurrent temperature profiles from a NOAA radio acoustic sounding system at McKinleyville (available at https://psl.noaa.gov/data/obs/datadisplay/).

b. Model configuration for 1-yr simulations

Based on a preliminary investigation of several aspects of the model configuration used for the CA20 dataset, we find that the choice of the PBL scheme used may be responsible for a large portion of the model bias (Bodini et al. 2022). Because the buoy deployment time frame was after the original 20-yr CA20 model output (i.e., from 2000 to 2019), a pair of 1-yr simulations with both the YSU and MYNN PBL schemes are conducted for a direct comparison with the floating lidars. These 1-yr simulations use the same model configuration as the CA20 dataset, except for the surface layer scheme (National Renewable Energy Laboratory 2023). As described by Optis et al. (2020), the Advanced Research version of the Weather Research and Forecasting (WRF-ARW) Model version 4.1.2 (Skamarock et al. 2019) is used. The model configuration consists of an outer domain (D01) with 6-km horizontal grid spacing and an inner domain (D02) with 2-km grid spacing, which roughly aligns with the domain setting shown in Fig. 1. The model includes 60 vertical layers, with a near 15–20-m interval for the lowest 200 m. The lateral boundary conditions for the outer domain are from the ERA5 reanalysis (Hersbach et al. 2020) with a spatial resolution of ∼30 km. The lateral boundary conditions for the inner domain are provided by the outer domain using one-way nesting. The model physics parameterization follows the CA20 model setup including the Ferrier (New Eta) microphysics scheme (Gallus and Segal 2001), the Noah land surface model (Chen and Dudhia 2001), and the Rapid Radiative Transfer Model longwave and shortwave radiation schemes (Iacono et al. 2008). The revised MM5 Monin–Obukhov surface layer scheme (Jiménez et al. 2012) is used for the 1-yr simulation to replace the MYNN surface layer scheme that was used in the original CA20 dataset. Simulations are run using both the MYNN and YSU PBL schemes and the results are compared. The WRF model simulations are conducted separately for each month with spectral nudging and then the data are concatenated into a single series. A spinup period of 2 days is used for each simulation. The wind speed is output every 5 min across the whole domain.

Fig. 1.
Fig. 1.

Simulation domain for the short-term simulations. The red dots show locations of the Humboldt and Morro Bay lidar buoys. The gray dots and orange squares show the locations of the NOAA surface buoys and onshore radar used to validate the CA20 dataset. The inserts show the locations of BOEM wind energy lease areas. The black line indicates the Humboldt–McKinleyville cross section.

Citation: Monthly Weather Review 152, 2; 10.1175/MWR-D-23-0030.1

c. Case selection and model configuration

Since the CA20 dataset does not archive all the variables needed for analyzing physical mechanisms driving the results, we conduct simulation over the entire July 2021 and a 3-day period in November 2020 when the wind speed and error are relatively large. Given the larger error at Humboldt compared to Morro Bay, our analyses will focus on the comparison with the Humboldt lidar buoy. The short-term simulations have a similar model setup compared to the 1-yr model output including model physics parameterizations and grid spacing but the domain covers a smaller region focusing on the area offshore California (Fig. 1). Another difference between the 1-yr simulations and the short-term simulations is the version of the WRF-ARW Model. The 1-yr simulations use WRF-ARW version 4.1.2, whereas the short-term simulations use WRF-ARW version 4.2 without nudging. Sensitivity analysis showed that the difference in results due to these modifications of simulation configuration had minimal effects on the conclusions of this study. For the short-term simulations, we perform WRF model simulations starting at 0000 UTC each day and lasting for 36 h. The results from the first 12 h are excluded as spinup, and the remaining 24 h are concatenated. This approach is different from starting the model monthly at 0000 UTC on the first day of each month and running for a continuous month, which was used for the CA20 model configuration. We expected the shorter integration period to help control error accumulation, but it showed a small impact on hub-height wind speeds in this study.

After evaluating the 1-yr CA20 WRF model simulations against observation, we observed larger errors during warm seasons, especially in July, while small errors in cool seasons. Therefore, we further investigate four cases from July 2021 and one case from November 2020 representing typical weather conditions during respective seasons. All five selected periods are associated with northerly winds but exhibit different background conditions. During 12–14 July, 21–23 July, and 18–20 July 2021, larger errors are found at Humboldt when using the MYNN PBL scheme compared to the YSU PBL scheme. Humboldt is mainly dominated by diurnally varying wind speeds and is underneath different offshore stratocumulus cloud cover conditions. This is a typical weather condition in warm months when the NPH dominates the offshore region, coinciding with the low pressure over the Southwest United States and the diurnally varying solar heating. Boundary layer clouds appear sometimes but are blocked from moving far onshore by the coastal mountains. During 5–7 July 2021, despite the persistence of the NPH, the models simulate coastal clouds that mitigate the diurnal varying land–sea thermal contrast and errors in wind speed. In winter, the passage of cold fronts also produces strong north winds and often with smaller biases compared to the periods with north winds in summer. We select a postfrontal case study during 7–9 November 2020, as an example of the typical wintertime weather condition.

3. Results

a. Prevailing wind regimes over the California outer continental shelf

Before diving into the analysis, we briefly summarize the prevailing wind regimes over the California outer continental shelf in summer and winter (Fig. 2). In summer, the NPH and thermal low over the Southwest U.S. result in an enhanced cross-coastline pressure gradient that primarily drives the prevailing northerly wind offshore (Beardsley et al. 1987; Zemba and Friehe 1987; Burk and Thompson 1996). The subsidence in NPH causes a thick air temperature inversion that is the lowest and strongest at the coast (Dorman et al. 2000). The subsidence inversion and the coastal mountains limit the MBL from the top and east side, respectively. At the top of the MBL, we typically observe a northerly to northwesterly low-level jet (LLJ) along the coastline that is primarily explained by thermal wind resulting from strong coastal baroclinity superimposed on the generally northerly flow (e.g., Burk and Thompson 1996; Fewings et al. 2016). The southerly thermal wind is manifested by free tropospheric winds that become more southerly away from the surface and more northerly toward the surface; thus, the wind speed increases toward the lower boundary until surface drag starts to affect it in the MBL at about 500 m. From there, wind speed decreases to the surface.

Fig. 2.
Fig. 2.

The LLJ and drivers over the California outer continental shelf.

Citation: Monthly Weather Review 152, 2; 10.1175/MWR-D-23-0030.1

The jet is located close to shore where the temperature gradient is greatest. The strength and location of the jet and the synoptic northerly flow are altered by many factors, including the land–sea thermal contrast and coastal terrain. North Californian coastal mountains are generally higher than the top of MBL, which forces the wind to follow the coastline. The major coastal capes, i.e., Cape Blanco and Cape Mendocino, significantly perturb the wind and temperature field within the MBL due to the hydraulic effect (Parish et al. 2016; Dorman et al. 1999), leading to slowed wind speed upwind of the cape and accelerated wind speed downwind (e.g., Edwards et al. 2001; Monteiro et al. 2016). These accelerated wind speeds result in local maxima within ∼15 km of the coast and on the downwind side of major capes, a phenomenon known as the expansion fan (Koračin and Dorman 2001; Dorman and Koračin 2008; Rahn et al. 2014), Fig. 3. Despite the similar synoptic weather conditions between Morro Bay and Humboldt, the wind speed at Morro Bay is less sensitive to simulated errors in the location of the low-level jet due to the lower coastal mountains and absence of capes and corresponding hydraulic effects and expansion fans. In winter, the California outer continental shelf has distinct synoptic-scale weather conditions. The NPH weakens in terms of central pressure and subsidence area (Schroeder et al. 2013), along with the deepening of the MBL (Lin et al. 2009). The weather conditions in the coastal area are associated with strong northerly winds due to storms and strong fronts from the Gulf of Alaska.

Fig. 3.
Fig. 3.

Mean wind speed simulated using the (a),(d) YSU and (b),(e) MYNN PBL schemes and (c),(f) the difference between the MYNN and YSU simulations during July 2021. Shown are (top) 100- and (bottom) 4-m wind speeds. The circles in (a), (b), (d), and (e) are colored with observed wind speed. The red dots in (c) and (f) indicate the locations of the Humboldt (north) and Morro Bay (south) lidar buoys. Note that the reference vectors are different in (c) and (f).

Citation: Monthly Weather Review 152, 2; 10.1175/MWR-D-23-0030.1

b. Errors in 1-yr simulation

By analyzing the available lidar data at Humboldt, we observed higher wind speed in warm seasons than cool seasons, with northward and southward prevailing wind direction. The 1-yr simulation with MYNN overestimates turbine hub-height wind speed at Humboldt by 1.9 m s−1 during the floating lidar observation period in contrast to the smaller bias of 0.6 m s−1 found for Morro Bay that is comparable to bias estimates from other regions (Hahmann et al. 2020). When binning data, positive biases are observed across all months (Fig. 4a) and at all hours of the day (not shown). The bias and large RMSE (Fig. 4b) are larger in warm months from June to October. We find significantly larger positive biases associated with the northerly wind component (Fig. 4c), which is the dominant condition in this region. The annual mean bias at Humboldt reduces to 0.13 m s−1 with YSU, with improvements in all the considered months. Furthermore, the model uncertainty is approximately 12% and 8% in Humboldt and Morro Bay, respectively, when using MYNN configurations while choosing YSU, the uncertainty reduces to 10% and 7%, at each site. Significant improvements are found during warm months with northerly wind and varying weather conditions.

Fig. 4.
Fig. 4.

Observed 100-m wind speed at Humboldt (a) bias and (b) RMSE as a function of month, and (c) bias/difference as a function of wind direction using the MYNN and YSU PBL schemes. Lines are for observed 100-m wind speed.

Citation: Monthly Weather Review 152, 2; 10.1175/MWR-D-23-0030.1

Over the entire July 2021, the MYNN simulation produces stronger expansion fans downwind the capes compared to those in YSU simulation (Fig. 3), leading to stronger 100-m wind speed at Humboldt. However, this is not observed at Morro Bay or from the near surface, consistent with the previous study (Optis et al. 2020).

Based on this analysis, we select four cases in July 2021 when large errors with the MYNN scheme are found in the 1-yr comparison and one case in November 2021 with small errors (shown in Table 1). Generally, similar synoptic weather systems are found throughout July 2021. Local thermodynamic disturbances (e.g., sea-breeze circulation) interacting with the offshore environment alter the wind speed near Humboldt, which results in different errors. Clear diurnal varying winds are observed during 12–15 July (case 1), 21–24 July (case 2), and 18–21 July (case 3) in 2021, along with large errors in the MYNN simulations (Table 1 and Fig. 5). The diurnal variation and errors are small during 5–8 July 2021 (case 4). In winter, although the northerly wind still dominates the offshore area during 7–10 November 2020 (case 5), the errors are small. In the following sections, we investigate the drivers of the errors by comparing model simulations and observations for each selected case study.

Table 1.

Errors and weather conditions for selected case studies. The statistics are obtained by comparing the simulated turbine hub-height (100 m) wind speed with concurrent lidar measurements at Humboldt. Numbers in the parentheses indicate the relative errors with respect to the mean wind speed during the case period (%). Model simulations are resampled to 10-min intervals. PWD stands for prevailing wind direction, and RMSE stands for root-mean-square error.

Table 1.
Fig. 5.
Fig. 5.

Time series of observed and simulated 100-m wind speed for (a) 12–15 Jul 2021, (b) 21–24 Jul 2021, (c) 18–21 Jul 2021, (d) 5–8 Jul 2021, and (e) 7–10 Nov 2020.

Citation: Monthly Weather Review 152, 2; 10.1175/MWR-D-23-0030.1

c. Large error cases in July 2021

During case 1 and case 2, the typical summertime weather system consists of the NPH and thermal low, driving synoptic northerly winds offshore (Fig. 6a). Expansion fans appear downwind of the Cape Blanco and Cape Mendocino at wind turbine heights. The Humboldt floating lidar is located between the two capes, where the MBL becomes deeper. Wind speed decreases sharply from the expansion fan downwind of Cape Blanco to slow winds upwind of Cape Mendocino. Large errors are found in the simulated 100-m wind speeds during these periods in the 1-yr simulation using the MYNN scheme with over 5 m s−1 RMSE compared to the Humboldt floating lidar (Table 1 and Fig. 6b). Much smaller errors are found in the YSU simulation, with less than 3 m s−1 RMSE. The mean wind speed in the MYNN simulation is over 3 m s−1 faster than observation and over 5 m s−1 compared to YSU. Positive differences widely spread offshore south of Oregon and in Northern California, with maximum values near Humboldt (Fig. 6c).

Fig. 6.
Fig. 6.

As in Fig. 3, but for 1800 UTC 3 Jul 2021 and only at 100 m.

Citation: Monthly Weather Review 152, 2; 10.1175/MWR-D-23-0030.1

The measured wind speed at Humboldt shows strong diurnal variation and vertical shear (Figs. 7a and 8a). At 100 m MSL, wind speed varies from 12 m s−1 during the daytime (2100–0300 UTC) to 22 m s−1 during the nighttime (0600–1200 UTC). The diurnal variation of the land–sea thermal contrast accounts for a large part of the observed diurnal variation in wind speed at the lidar location via the thermal wind theory. A detailed discussion about the diurnal variation in wind speed and sea-breeze circulation will be given later. Vertically, observed wind speed increases sharply from near surface to 240 m MSL, which is captured by both parameterizations while neither model matches the magnitude of the observed winds themselves (Figs. 7b,c). While the MYNN simulation overestimates the daytime wind speed by a larger magnitude than at night, the diurnal variation is muted (Figs. 5a,b and 7c). Figures 7a–c compare the vertical structure below 240 m due to the height limit of floating lidar measurement. On the other hand, Fig. 7d shows the difference between the MYNN and YSU simulations from the surface up to 1000 m. The MYNN simulation produces larger wind speed only below 400 m, with a maximum difference over 10 m s−1 in the early afternoon. Above 400 m, the wind speed from the MYNN simulation is actually slower than what is modeled by the YSU scheme.

Fig. 7.
Fig. 7.

Time–height cross sections of wind speed at Humboldt from the (a) lidar, (b) YSU simulation, (c) MYNN simulation, and (d) MYNN minus YSU during 12–15 Jul 2021 (note the different extent of the vertical axis). The blue curve in (a) indicates the temperature difference between McKinleyville and Humboldt at 300 m.

Citation: Monthly Weather Review 152, 2; 10.1175/MWR-D-23-0030.1

Fig. 8.
Fig. 8.

Vertical profiles of wind speed at Humboldt at 1200 UTC (solid line) and 0000 UTC (dashed line) for (a) 12–15 Jul 2021, (b) 21–24 Jul 2021, (c) 18–21 Jul 2021, (d) 5–8 Jul 2021, and (e) 7–10 Nov 2020.

Citation: Monthly Weather Review 152, 2; 10.1175/MWR-D-23-0030.1

To understand the drivers of the positive biases below 400 m found in the MYNN simulation, we analyze the Humboldt–McKinleyville east-west cross-sections of temperature and wind speed, as shown in Fig. 9. The subsidence of the NPH creates an inversion layer near an altitude of 500 m, which limits the top height of the MBL. In the early afternoon, the mixed layer temperature inland is much warmer than the MBL. At night, the inland minim surface temperature is slightly smaller than the adjacent sea surface temperature (SST). Above the shallow surface inversion that develops over land, the inland temperature remains larger at all times than at comparable levels over the adjacent ocean. The land–sea thermal contrast creates a strong baroclinic zone that primarily drives the LLJ (Figs. 9a,c). The simulated jet core, with wind speed greater than 20 m s−1, is 200–800 m above the ocean west of the Humboldt floating lidar. The MBL becomes deeper at Humboldt due to the presence of Cape Mendocino on the downwind side. From the cross-section view, the jet becomes thinner closer to the coast. The Humboldt buoy sits in the low wind speed MBL under the jet where the vertical shear is large (Figs. 7a and 8a). At the turbine height, the Humboldt floating lidar is located at the edge of the expansion fan caused by Cape Blanco (Fig. 6a). Wind speed at Humboldt, therefore, is sensitive to perturbations due to the sharp gradient of wind speed both vertically and horizontally.

Fig. 9.
Fig. 9.

Humboldt–McKinleyville cross section (shown in Fig. 1) of (a) temperature (°C) and zonal and vertical wind vector (m s−1) from YSU simulation, and (b) temperature difference between MYNN and YSU at 1800 UTC 13 Jul 2021 during case 1. (c),(d) As in (a) and (b), but for V wind (m s−1). The red dots show the location of the radar wind profiler near the coast and the offshore buoy at Humboldt. The vertical wind speed is scaled by multiplying 100.

Citation: Monthly Weather Review 152, 2; 10.1175/MWR-D-23-0030.1

The diurnal variation of wind speed at Humboldt is in large part due to the sea-breeze circulation induced by the land–sea thermal contrast. As discussed by Burk and Thompson (1996), in the daytime, when the sea-breeze circulation intensifies, the jet core lowers and moves toward the coast. During nighttime, when the strong sea-breeze circulation dissipates, the jet core lifts and moves away from the coast. The jet movement is associated with the MBL dynamics. The coastal MBL lowers in the afternoon and deepens at night in response to the enhanced and weakened subsidence over coastal waters induced by the sea-breeze circulation, respectively. The jet follows the strong horizontal temperature gradient at the boundary layer top as a result of the thermal wind forcing but with a temporal lag of a few hours. The maximum wind speed of the jet occurs around 0500 UTC, which is later than the peak of baroclinity around 2300 UTC (Fig. 7a). The discrepancy is likely due to the inertial turning of the onshore sea breeze circulation from the day, which takes a few hours but rotates the wind direction to northerly, which adds to the thermal wind component.

With the MYNN scheme, the model-simulated land surface temperature over the coastal mountains is 3°C warmer than the YSU-simulated one throughout the day, whereas the near sea surface temperature is similar in the two models (Fig. 7b). The temperature difference is larger in the midafternoon, with a 5°C warmer temperature in the MYNN simulations. The additional sea-breeze circulation associated with the larger temperature difference increases the subsidence near the coast and the low-level divergence in the coastal MBL, leading to a thinner MBL in the MYNN simulations. Consequently, the jet core lowers and moves toward the coast, which increases the turbine hub-height wind speed in the afternoon (a few hours lag from the maximum intensification of the sea-breeze circulation). The resultant wind speed acceleration is stronger in the daytime when the wind speed is expected to be small (Figs. 5a,b). Therefore, the diurnal variation of wind speed is muted in the MYNN simulation.

d. Moderate and small error cases in July 2021

Moderate wind speed errors are found in the MYNN simulation during 18–21 July 2021 (case 3), although the influencing synoptic weather systems are similar to cases 1 and 2. With the MYNN scheme, the model overestimates the 100-m wind speed by 3.2 m s−1, with RMSE 3.67 m s−1, whereas the bias and RMSE with YSU scheme are 0.32 and 1.77 m s−1, respectively (Table 1). The diurnal variation is well captured by both models (Fig. 5c).

Besides the same mechanism that causes large errors in cases 1 and 2, we find synoptic-scale westerly flows during case 3. Those west winds bring moist and cool marine air inland, cooling the coast and reducing the baroclinicity. The sea-breeze circulation is weakened, and the corresponding subsidence branch is constrained close to the coastline. Consequently, the jet core is weakened and moves farther offshore (Fig. 10c). Although with MYNN, the model still simulates moderately land–sea thermal contrast, the overestimation and associated impact in the wind speed is much smaller.

Fig. 10.
Fig. 10.

As in Fig. 9, but for 1800 UTC 19 Jul 2021.

Citation: Monthly Weather Review 152, 2; 10.1175/MWR-D-23-0030.1

Errors are further decreased during 5–8 July 2021 (case 4). The synoptic-scale environment is altered by strong westerly flows associated with MBL clouds propagating onshore. Onshore winds efficiently cool the coast, breaking down the land–sea thermal contrast. This breakdown causes the baroclinity to be lessened considerably and thereby diminishing the jet structure (Fig. 11c). The thermal-driven diurnal variation in wind speed is also muted (Fig. 5d). Meanwhile, the marine air intrusion dominates the coastal mountains, resulting in a small difference in air temperature between the two simulations. Combining these two aspects, the model with the MYNN scheme produces a slightly smaller RMSE and similar absolute bias than those with the YSU scheme (Fig. 8d).

Fig. 11.
Fig. 11.

As in Fig. 9, but for 1800 UTC 6 Jul 2021.

Citation: Monthly Weather Review 152, 2; 10.1175/MWR-D-23-0030.1

e. Small error case in November 2020

The wintertime California outer continental shelf is often dominated by storms and strong fronts originating in the Gulf of Alaska that bring strong northerly winds to the study area. Errors in both MYNN and YSU simulations in winter are considerably smaller than in summer (Figs. 4a and 8e). A cold front passes Humboldt on 6 November 2020, leaving behind strong northerly to northwesterly winds during 7–10 November 2020 (case 5). Due to the lack of the thermal-wind mechanism, the synoptic-scale forcing is mainly responsible for strong northerly winds in the lower atmosphere (Fig. 12c). The cold air passes over the warmer SST leading to a typical well-mixed wintertime MBL. Vertical wind shear is smaller than those with LLJs in summer.

Fig. 12.
Fig. 12.

As in Fig. 9, but for 1800 UTC 7 Nov 2020.

Citation: Monthly Weather Review 152, 2; 10.1175/MWR-D-23-0030.1

Models with the two PBL schemes simulate similar synoptic-scale thermal conditions, whereas the model setup with MYNN produces faster winds than with YSU, especially above 500 m (Figs. 12b,d). At the wind turbine height, the difference in wind speed is negligible (Table 1).

4. Discussion and conclusions

In this study, we investigate the large wind speed errors found in the CA20 dataset at Humboldt compared to the recently available lidar buoy measurements. Comparing 1-yr simulations with the MYNN and YSU PBL schemes, we find that the errors have a strong correlation with wind direction regimes and seasons: errors are generally larger with the MYNN scheme during warm seasons with northerly winds. To better understand the mechanisms that can contribute to the observed bias, we conduct five short-term simulations using the two PBL schemes. In the summer, the NPH and thermal low over the Southwest United States primarily drive equatorward wind parallel to the coastline. Except for nighttime near the surface, air temperature over the coastal mountains is warmer than over the adjacent ocean at the same height throughout the day (Fig. 14a). This land–sea thermal contrast accelerates the northerly winds superimposed on the generally northerly flow via the thermal-wind mechanism. The resultant low-level jet is often observed along the coast. The diurnal variation of the sea-breeze forcing alters the location and strength of the jet, which leads to the diurnally varying wind speed regime observed at Humboldt.

In MYNN simulations, our case studies show a warm air temperature bias over the coastal mountains (Fig. 13), which accelerates the northerly winds further, especially in the morning to early afternoon. Consequently, the diurnal variation in wind speed is muted, and a large positive bias in MYNN-simulated wind speed is observed (Fig. 14b). When synoptic-scale winds have a westerly component, the marine air intrusion cools the coastal mountains, diminishing the see-breeze circulation, which remains confined close to the coast. As a consequence, the thermally induced acceleration moves further offshore, having a smaller impact at Humboldt. Therefore, we find that the associated difference between MYNN and YSU is small during cases with such weather conditions.

Fig. 13.
Fig. 13.

Time series of virtual temperature at McKinleyville at 300 m in July. The shaded regions show different time periods corresponding to the various case studies analyzed.

Citation: Monthly Weather Review 152, 2; 10.1175/MWR-D-23-0030.1

Fig. 14.
Fig. 14.

Drivers associated with the LLJ in the (a) YSU and (b) MYNN simulations. The differences between the two simulations are marked in red in (b).

Citation: Monthly Weather Review 152, 2; 10.1175/MWR-D-23-0030.1

The warm bias associated with the MYNN PBL scheme has been reported in previous studies. In particular, the operational implementation of the High-Resolution Rapid Refresh (HRRR) numerical weather prediction system at NCEP has highlighted this bias since its initial version in 2014 (James et al. 2022). The warm bias is attributed to insufficient low-level cloud cover, leading to an increased downward shortwave radiation. The developments both in data assimilation and model physics alleviate the warm bias in the latest HRRRv4; however, the warm bias still exists. In our case, we find that the model with MYNN scheme simulated a smaller low-level cloud fraction along the coast and offshore in July 2021 (Fig. 15). The reduction in cloud cover results in an increased downward shortwave, leading to warmer surface. The SST is prescribed. Therefore, the warm bias in MYNN simulation amplifies the land–sea thermal gradient, consequently, enhances the low-level jet. Besides, we also note that mixing length factors in MYNN could play an important role in causing the warm bias (Olson et al. 2019b).

Fig. 15.
Fig. 15.

Differences in (a) cloud fraction, (b) downward shortwave, and (c) 2-m temperature using simulations with MYNN minus YSU, averaged over July 2021.

Citation: Monthly Weather Review 152, 2; 10.1175/MWR-D-23-0030.1

In the summer, wind-driven coastal upwelling brings cold water to the surface, causing a cold pool off the coast (Figs. 911). Colder sea surface sits underneath warm air, leading to a stable MBL. The stable MBL suppresses vertical mixing and produces strong wind shear (Optis et al. 2020). Using different SST datasets could alter the stability of MBL and is anticipated to affect boundary layer wind speeds. The effect of different SST products, such as Coupled Ocean/Atmosphere Mesoscale Prediction System (Hodur et al. 2002), Operational Sea Surface Temperature and Ice Analysis (Donlon et al. 2012), and National Centers for Environmental Prediction (Saha et al. 2010) as bottom boundary conditions do not affect our results from this study [also see Fig. 9 in Optis et al. (2020)]. Burk and Thompson (1996) warmed up the cold pool by broadcasting the west boundary SST. They found that in the west side of the jet core the momentum is somewhat more well-mixed vertically than in their control simulation. However, this difference is not sufficiently strong to have a major impact on boundary layer structure and, thereby, the boundary layer wind speed.

The CA20 WRF Model setup was selected based on physical consideration after validating with an array of California near-surface buoys (generally at 4 m above sea surface level), at three California coastal radars (195 m above ground level), and at two floating lidars in the mid-Atlantic. Due to the strong vertical wind shear and land–sea gradient, those larger errors at higher heights were not detected until comparing against floating lidar measurements that have recently become available. Our findings suggest having access to observed hub-height wind speed is essential to ensure an accurate model validation, which is also suggested by Draxl et al. (2012). We also highlight the validation of other wind-related variables such as surface temperature where coastal thermal circulations could matter. By leveraging the physical knowledge from this study and companion efforts, the team will build a 20-yr updated version of the CA20 dataset, which is expected to be publicly available in spring 2023.

Acknowledgments.

The authors thank Colleen M. Kaul from PNNL for reviewing earlier versions of the manuscript. The Pacific Northwest National Laboratory is operated by DOE by the Battelle Memorial Institute under Contract DE-A05-76RL0 1830. This work was authored in part by the National Renewable Energy Laboratory operated by Alliance for Sustainable Energy, LLC, for the U.S. Department of Energy (DOE) under Contract DE-AC36-08GO28308. Funding provided by the U.S. Department of Energy Office of Energy Efficiency and Renewable Energy Wind Energy Technologies Office. The views expressed in the article do not necessarily represent the views of the DOE or the U.S. government. The U.S. government retains and the publisher, by accepting the article for publication, acknowledges that the U.S. government retains a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this work, or allow others to do so, for U.S. government purposes.

Data availability statement.

The lidar buoy data are publicly available at https://a2e.energy.gov/about/dap. Near-surface, wave, current, and cloud datasets for Humboldt and Morro Bay buoys are provided at https://doi.org/10.21947/1783807 and https://doi.org/10.21947/1959715, respectively. Lidar datasets for Humboldt and Morro Bay are provided at https://doi.org/10.21947/1783809 and https://doi.org/10.21947/1959721, respectively. The numerical simulations are too large to archive or to transfer. The information needed to replicate the simulation are available from https://data.openei.org/files/4500/WRF_NamelistFiles_NOW-23_v2.zip. Apart from the selection of MYNN and YSU as the PBL scheme, the only disparity in the namelist is that the original CA20 simulation used the MYNN surface layer scheme, whereas the new simulations use the Revised MM5 scheme.

REFERENCES

  • Beardsley, R. C., C. E. Dorman, C. A. Friehe, L. K. Rosenfeld, and C. D. Winant, 1987: Local atmospheric forcing during the coastal ocean dynamics experiment: 1. A description of the marine boundary layer and atmospheric conditions over a northern California upwelling region. J. Geophys. Res., 92, 14671488, https://doi.org/10.1029/JC092iC02p01467.

    • Search Google Scholar
    • Export Citation
  • Berg, L. K., Y. Liu, B. Yang, Y. Qian, J. Olson, M. Pekour, P.-L. Ma, and Z. Hou, 2019: Sensitivity of turbine-height wind speeds to parameters in the planetary boundary-layer parametrization used in the Weather Research and Forecasting Model: Extension to wintertime conditions. Bound.-Layer Meteor., 170, 507518, https://doi.org/10.1007/s10546-018-0406-y.

    • Search Google Scholar
    • Export Citation
  • Berg, L. K., Y. Liu, B. Yang, Y. Qian, R. Krishnamurthy, L. Sheridan, and J. Olson, 2021: Time evolution and diurnal variability of the parametric sensitivity of turbine‐height winds in the MYNN‐EDMF parameterization. J. Geophys. Res. Atmos., 126, e2020JD034000, https://doi.org/10.1029/2020JD034000.

    • Search Google Scholar
    • Export Citation
  • Bodini, N., and Coauthors, 2022: Update on NREL’s 2020 offshore wind resource assessment for the California Pacific outer continental shelf. NREL Rep. NREL/TP-5000-83756, 25 pp., https://www.osti.gov/biblio/1899984.

  • Burk, S. D., and W. T. Thompson, 1996: The summertime low-level jet and marine boundary layer structure along the California coast. Mon. Wea. Rev., 124, 668686, https://doi.org/10.1175/1520-0493(1996)124<0668:TSLLJA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Chen, F., and J. Dudhia, 2001: Coupling an advanced land surface–hydrology model with the Penn State–NCAR MM5 modeling system. Part I: Model implementation and sensitivity. Mon. Wea. Rev., 129, 569585, https://doi.org/10.1175/1520-0493(2001)129<0569:CAALSH>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Deardorff, J. W., 1972: Parameterization of the planetary boundary layer for use in general circulation models. Mon. Wea. Rev., 100, 93106, https://doi.org/10.1175/1520-0493(1972)100<0093:POTPBL>2.3.CO;2.

    • Search Google Scholar
    • Export Citation
  • Dettinger, M. D., F. M. Ralph, T. Das, P. J. Neiman, and D. R. Cayan, 2011: Atmospheric rivers, floods and the water resources of California. Water, 3, 445478, https://doi.org/10.3390/w3020445.

    • Search Google Scholar
    • Export Citation
  • Donlon, C. J., M. Martin, J. Stark, J. Roberts-Jones, E. Fiedler, and W. Wimmer, 2012: The Operational Sea Surface Temperature and Sea Ice Analysis (OSTIA) system. Remote Sens. Environ., 116, 140158, https://doi.org/10.1016/j.rse.2010.10.017.

    • Search Google Scholar
    • Export Citation
  • Dorman, C. E., and D. Koračin, 2008: Response of the summer marine layer flow to an extreme California coastal bend. Mon. Wea. Rev., 136, 28942922, https://doi.org/10.1175/2007MWR2336.1.

    • Search Google Scholar
    • Export Citation
  • Dorman, C. E., D. P. Rogers, W. Nuss, and W. T. Thompson, 1999: Adjustment of the summer marine boundary layer around Point Sur, California. Mon. Wea. Rev., 127, 21432159, https://doi.org/10.1175/1520-0493(1999)127<2143:AOTSMB>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Dorman, C. E., T. Holt, D. P. Rogers, and K. Edwards, 2000: Large-scale structure of the June–July 1996 marine boundary layer along California and Oregon. Mon. Wea. Rev., 128, 16321652, https://doi.org/10.1175/1520-0493(2000)128<1632:LSSOTJ>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Draxl, C., A. N. Hahmann, A. Peña, and G. Giebel, 2012: Evaluating winds and vertical wind shear from weather research and forecasting model forecasts using seven planetary boundary layer schemes. Wind Energy, 17, 3955, https://doi.org/10.1002/we.1555.

    • Search Google Scholar
    • Export Citation
  • Draxl, C., A. Clifton, B.-M. Hodge, and J. McCaa, 2015: The Wind Integration National Dataset (WIND) toolkit. Appl. Energy, 151, 355366, https://doi.org/10.1016/j.apenergy.2015.03.121.

    • Search Google Scholar
    • Export Citation
  • Edwards, K. A., A. M. Rogerson, C. D. Winant, and D. P. Rogers, 2001: Adjustment of the marine atmospheric boundary layer to a coastal CAPE. J. Atmos. Sci., 58, 15111528, https://doi.org/10.1175/1520-0469(2001)058<1511:AOTMAB>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Fewings, M. R., L. Washburn, C. E. Dorman, C. Gotschalk, and K. Lombardo, 2016: Synoptic forcing of wind relaxations at Pt. Conception, California. J. Geophys. Res. Oceans, 121, 57115730, https://doi.org/10.1002/2016JC011699.

    • Search Google Scholar
    • Export Citation
  • Guan, B., and D. E. Waliser, 2015: Detection of atmospheric rivers: Evaluation and application of an algorithm for global studies. J. Geophys. Res. Atmos., 120, 12 51412 535, https://doi.org/10.1002/2015JD024257.

    • Search Google Scholar
    • Export Citation
  • Hahmann, A. N., and Coauthors, 2020: The making of the new European wind atlas—Part 1: Model sensitivity. Geosci. Model Dev., 13, 50535078, https://doi.org/10.5194/gmd-13-5053-2020.

    • Search Google Scholar
    • Export Citation
  • Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 19992049, https://doi.org/10.1002/qj.3803.

    • Search Google Scholar
    • Export Citation
  • Hodur, R. M., X. Hong, J. D. Doyle, J. Pullen, J. Cummings, P. Martin, and M. A. Rennick, 2002: The Coupled Ocean/Atmosphere Mesoscale Prediction System (COAMPS). Oceanography, 15, 8898, https://doi.org/10.5670/oceanog.2002.39.

    • Search Google Scholar
    • Export Citation
  • Hong, S.-Y., Y. Noh, and J. Dudhia, 2006: A new vertical diffusion package with an explicit treatment of entrainment processes. Mon. Wea. Rev., 134, 23182341, https://doi.org/10.1175/MWR3199.1.

    • Search Google Scholar
    • Export Citation
  • Iacono, M. J., J. S. Delamere, E. J. Mlawer, M. W. Shephard, S. A. Clough, and W. D. Collins, 2008: Radiative forcing by long‐lived greenhouse gases: Calculations with the AER radiative transfer models. J. Geophys. Res., 113, D13103, https://doi.org/10.1029/2008JD009944.

    • Search Google Scholar
    • Export Citation
  • James, E. P., and Coauthors, 2022: The High-Resolution Rapid Refresh (HRRR): An hourly updating convection-allowing forecast model. Part II: Forecast performance. Wea. Forecasting, 37, 13971417, https://doi.org/10.1175/WAF-D-21-0130.1.

    • Search Google Scholar
    • Export Citation
  • Jiménez, P. A., J. Dudhia, J. F. González-Rouco, J. Navarro, J. P. Montávez, and E. García-Bustamante, 2012: A revised scheme for the WRF surface layer formulation. Mon. Wea. Rev., 140, 898918, https://doi.org/10.1175/MWR-D-11-00056.1.

    • Search Google Scholar
    • Export Citation
  • Gallus, W. A., Jr., and M. Segal, 2001: Impact of improved initialization of mesoscale features on convective system rainfall in 10-km eta simulations. Wea. Forecasting, 16, 680696, https://doi.org/10.1175/1520-0434(2001)016<0680:IOIIOM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Koračin, D., and C. E. Dorman, 2001: Marine atmospheric boundary layer divergence and clouds along California in June 1996. Mon. Wea. Rev., 129, 20402056, https://doi.org/10.1175/1520-0493(2001)129<2040:MABLDA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Krishnamurthy, R., and L. Sheridan, 2020a: California–Wind Sentinel (120), Humboldt/reviewed data. U.S. Department of Energy, accessed 1 May 2023, https://doi.org/10.21947/1783807.

  • Krishnamurthy, R., and L. Sheridan, 2020b: California–Wind Sentinel (130), Morro Bay/reviewed data. U.S. Department of Energy, accessed 1 May 2023, https://doi.org/10.21947/1959715.

  • Krishnamurthy, R., and L. Sheridan, 2021a: California–Leosphere Windcube 866 (120), Humboldt/reviewed data. U.S. Department of Energy, accessed 1 May 2023, https://doi.org/10.21947/1783809.

  • Krishnamurthy, R., and L. Sheridan, 2021b: California–Leosphere Windcube 866 (130), Morro Bay/reviewed data. U.S. Department of Energy, accessed 1 May 2023, https://doi.org/10.21947/1959721.

  • Krishnamurthy, R., and Coauthors, 2023: Year-long buoy-based observations of the air–sea transition zone off the U.S. West Coast. Earth Syst. Sci. Data, 15, 56675699, https://doi.org/10.5194/essd-15-5667-2023.

    • Search Google Scholar
    • Export Citation
  • Lin, W., M. Zhang, and N. G. Loeb, 2009: Seasonal variation of the physical properties of marine boundary layer clouds off the California coast. J. Climate, 22, 26242638, https://doi.org/10.1175/2008jcli2478.1.

    • Search Google Scholar
    • Export Citation
  • Liu, Y., Y. Qian, and L. K. Berg, 2022: Local-thermal-gradient and large-scale-circulation impacts on turbine-height wind speed forecasting over the Columbia River basin. Wind Energy Sci., 7, 3751, https://doi.org/10.5194/wes-7-37-2022.

    • Search Google Scholar
    • Export Citation
  • Monteiro, I. T., A. J. Santos, M. Belo‐Pereira, and P. B. Oliveira, 2016: Adjustment of the summertime marine atmospheric boundary layer to the western Iberia coastal morphology. J. Geophys. Res. Atmos., 121, 38753893, https://doi.org/10.1002/2016JD025055.

    • Search Google Scholar
    • Export Citation
  • Musial, W., P. Spitsen, P. Duffy, P. Beiter, M. Marquis, R. Hammond, and M. Shields, 2022: Offshore wind market report: 2022 edition. NREL Rep., 126 pp., https://www.energy.gov/sites/default/files/2022-08/offshore_wind_market_report_2022.pdf.

  • Nakanishi, M., and H. Niino, 2009: Development of an improved turbulence closure model for the atmospheric boundary layer. J. Meteor. Soc. Japan, 87, 895912, https://doi.org/10.2151/jmsj.87.895.

    • Search Google Scholar
    • Export Citation
  • National Renewable Energy Laboratory, 2023: 2023 National Offshore Wind data set (NOW-23). NREL, accessed 23 October 2023, https://doi.org/10.25984/1821404.

  • Newsom, R., and R. K. Krishnamurthy, 2022: Doppler lidar (DL) instrument handbook. Rep. DOE/SC-ARM-TR-101, 59 pp., https://doi.org/10.2172/1034640.

  • Olson, J. B., and Coauthors, 2019a: Improving wind energy forecasting through numerical weather prediction model development. Bull. Amer. Meteor. Soc., 100, 22012220, https://doi.org/10.1175/BAMS-D-18-0040.1.

    • Search Google Scholar
    • Export Citation
  • Olson, J. B., J. S. Kenyon, W. Angevine, J. M. Brown, M. Pagowski, and K. Sušelj, 2019b: A description of the MYNN-EDMF scheme and the coupling to other components in WRF–ARW. NOAA Tech. Memo. OAR GSD-61, 42 pp., https://repository.library.noaa.gov/view/noaa/19837.

  • Optis, M., O. Rybchuk, N. Bodini, M. Rossol, and W. Musial, 2020: Offshore wind resource assessment for the California Pacific outer continental shelf (2020). NREL Rep. NREL/TP-5000-77642, 61 pp., https://doi.org/10.2172/1677466.

  • Parish, T. R., D. A. Rahn, and D. C. Leon, 2016: Aircraft measurements and numerical simulations of an expansion fan off the California coast. J. Appl. Meteor. Climatol., 55, 20532062, https://doi.org/10.1175/JAMC-D-16-0101.1.

    • Search Google Scholar
    • Export Citation
  • Payne, A. E., and Coauthors, 2020: Responses and impacts of atmospheric rivers to climate change. Nat. Rev. Earth Environ., 1, 143157, https://doi.org/10.1038/s43017-020-0030-5.

    • Search Google Scholar
    • Export Citation
  • Pronk, V., N. Bodini, M. Optis, J. K. Lundquist, P. Moriarty, C. Draxl, A. Purkayastha, and E. Young, 2022: Can reanalysis products outperform mesoscale numerical weather prediction models in modeling the wind resource in simple terrain? Wind Energy Sci., 7, 487504, https://doi.org/10.5194/wes-7-487-2022.

    • Search Google Scholar
    • Export Citation
  • Rahn, D. A., T. R. Parish, and D. Leon, 2014: Coastal jet adjustment near Point Conception, California, with opposing wind in the bight. Mon. Wea. Rev., 142, 13441360, https://doi.org/10.1175/MWR-D-13-00177.1.

    • Search Google Scholar
    • Export Citation
  • Saha, S., and Coauthors, 2010: The NCEP Climate Forecast System Reanalysis. Bull. Amer. Meteor. Soc., 91, 10151058, https://doi.org/10.1175/2010BAMS3001.1.

    • Search Google Scholar
    • Export Citation
  • Schroeder, I. D., B. A. Black, W. J. Sydeman, S. J. Bograd, E. L. Hazen, J. A. Santora, and B. K. Wells, 2013: The North Pacific High and wintertime pre‐conditioning of California current productivity. Geophys. Res. Lett., 40, 541546, https://doi.org/10.1002/grl.50100.

    • Search Google Scholar
    • Export Citation
  • Skamarock, W. C., and Coauthors, 2019: A description of the Advanced Research WRF Model version 4. NCAR Tech. Note NCAR/TN-556+STR, 145 pp., https://doi.org/10.5065/1dfh-6p97.

  • Smith, C., B. Hatchett, and M. Kaplan, 2018: Characteristics of sundowner winds near Santa Barbara, CA, from a dynamically downscaled climatology: Environment and effects aloft and offshore. J. Geophys. Res. Atmos., 123, 13 09213 110, https://doi.org/10.1029/2018JD029065.

    • Search Google Scholar
    • Export Citation
  • Soares, P. M. M., P. M. A. Miranda, A. P. Siebesma, and J. Teixeira, 2004: An eddy‐diffusivity/mass‐flux parametrization for dry and shallow cumulus convection. Quart. J. Roy. Meteor. Soc., 130, 33653383, https://doi.org/10.1256/qj.03.223.

    • Search Google Scholar
    • Export Citation
  • Zemba, J., and C. A. Friehe, 1987: The marine atmospheric boundary layer jet in the Coastal Ocean Dynamics Experiment. J. Geophys. Res., 92, 14891496, https://doi.org/10.1029/JC092iC02p01489.

    • Search Google Scholar
    • Export Citation
Save
  • Beardsley, R. C., C. E. Dorman, C. A. Friehe, L. K. Rosenfeld, and C. D. Winant, 1987: Local atmospheric forcing during the coastal ocean dynamics experiment: 1. A description of the marine boundary layer and atmospheric conditions over a northern California upwelling region. J. Geophys. Res., 92, 14671488, https://doi.org/10.1029/JC092iC02p01467.

    • Search Google Scholar
    • Export Citation
  • Berg, L. K., Y. Liu, B. Yang, Y. Qian, J. Olson, M. Pekour, P.-L. Ma, and Z. Hou, 2019: Sensitivity of turbine-height wind speeds to parameters in the planetary boundary-layer parametrization used in the Weather Research and Forecasting Model: Extension to wintertime conditions. Bound.-Layer Meteor., 170, 507518, https://doi.org/10.1007/s10546-018-0406-y.

    • Search Google Scholar
    • Export Citation
  • Berg, L. K., Y. Liu, B. Yang, Y. Qian, R. Krishnamurthy, L. Sheridan, and J. Olson, 2021: Time evolution and diurnal variability of the parametric sensitivity of turbine‐height winds in the MYNN‐EDMF parameterization. J. Geophys. Res. Atmos., 126, e2020JD034000, https://doi.org/10.1029/2020JD034000.

    • Search Google Scholar
    • Export Citation
  • Bodini, N., and Coauthors, 2022: Update on NREL’s 2020 offshore wind resource assessment for the California Pacific outer continental shelf. NREL Rep. NREL/TP-5000-83756, 25 pp., https://www.osti.gov/biblio/1899984.

  • Burk, S. D., and W. T. Thompson, 1996: The summertime low-level jet and marine boundary layer structure along the California coast. Mon. Wea. Rev., 124, 668686, https://doi.org/10.1175/1520-0493(1996)124<0668:TSLLJA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Chen, F., and J. Dudhia, 2001: Coupling an advanced land surface–hydrology model with the Penn State–NCAR MM5 modeling system. Part I: Model implementation and sensitivity. Mon. Wea. Rev., 129, 569585, https://doi.org/10.1175/1520-0493(2001)129<0569:CAALSH>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Deardorff, J. W., 1972: Parameterization of the planetary boundary layer for use in general circulation models. Mon. Wea. Rev., 100, 93106, https://doi.org/10.1175/1520-0493(1972)100<0093:POTPBL>2.3.CO;2.

    • Search Google Scholar
    • Export Citation
  • Dettinger, M. D., F. M. Ralph, T. Das, P. J. Neiman, and D. R. Cayan, 2011: Atmospheric rivers, floods and the water resources of California. Water, 3, 445478, https://doi.org/10.3390/w3020445.

    • Search Google Scholar
    • Export Citation
  • Donlon, C. J., M. Martin, J. Stark, J. Roberts-Jones, E. Fiedler, and W. Wimmer, 2012: The Operational Sea Surface Temperature and Sea Ice Analysis (OSTIA) system. Remote Sens. Environ., 116, 140158, https://doi.org/10.1016/j.rse.2010.10.017.

    • Search Google Scholar
    • Export Citation
  • Dorman, C. E., and D. Koračin, 2008: Response of the summer marine layer flow to an extreme California coastal bend. Mon. Wea. Rev., 136, 28942922, https://doi.org/10.1175/2007MWR2336.1.

    • Search Google Scholar
    • Export Citation
  • Dorman, C. E., D. P. Rogers, W. Nuss, and W. T. Thompson, 1999: Adjustment of the summer marine boundary layer around Point Sur, California. Mon. Wea. Rev., 127, 21432159, https://doi.org/10.1175/1520-0493(1999)127<2143:AOTSMB>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Dorman, C. E., T. Holt, D. P. Rogers, and K. Edwards, 2000: Large-scale structure of the June–July 1996 marine boundary layer along California and Oregon. Mon. Wea. Rev., 128, 16321652, https://doi.org/10.1175/1520-0493(2000)128<1632:LSSOTJ>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Draxl, C., A. N. Hahmann, A. Peña, and G. Giebel, 2012: Evaluating winds and vertical wind shear from weather research and forecasting model forecasts using seven planetary boundary layer schemes. Wind Energy, 17, 3955, https://doi.org/10.1002/we.1555.

    • Search Google Scholar
    • Export Citation
  • Draxl, C., A. Clifton, B.-M. Hodge, and J. McCaa, 2015: The Wind Integration National Dataset (WIND) toolkit. Appl. Energy, 151, 355366, https://doi.org/10.1016/j.apenergy.2015.03.121.

    • Search Google Scholar
    • Export Citation
  • Edwards, K. A., A. M. Rogerson, C. D. Winant, and D. P. Rogers, 2001: Adjustment of the marine atmospheric boundary layer to a coastal CAPE. J. Atmos. Sci., 58, 15111528, https://doi.org/10.1175/1520-0469(2001)058<1511:AOTMAB>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Fewings, M. R., L. Washburn, C. E. Dorman, C. Gotschalk, and K. Lombardo, 2016: Synoptic forcing of wind relaxations at Pt. Conception, California. J. Geophys. Res. Oceans, 121, 57115730, https://doi.org/10.1002/2016JC011699.

    • Search Google Scholar
    • Export Citation
  • Guan, B., and D. E. Waliser, 2015: Detection of atmospheric rivers: Evaluation and application of an algorithm for global studies. J. Geophys. Res. Atmos., 120, 12 51412 535, https://doi.org/10.1002/2015JD024257.

    • Search Google Scholar
    • Export Citation
  • Hahmann, A. N., and Coauthors, 2020: The making of the new European wind atlas—Part 1: Model sensitivity. Geosci. Model Dev., 13, 50535078, https://doi.org/10.5194/gmd-13-5053-2020.

    • Search Google Scholar
    • Export Citation
  • Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 19992049, https://doi.org/10.1002/qj.3803.

    • Search Google Scholar
    • Export Citation
  • Hodur, R. M., X. Hong, J. D. Doyle, J. Pullen, J. Cummings, P. Martin, and M. A. Rennick, 2002: The Coupled Ocean/Atmosphere Mesoscale Prediction System (COAMPS). Oceanography, 15, 8898, https://doi.org/10.5670/oceanog.2002.39.

    • Search Google Scholar
    • Export Citation
  • Hong, S.-Y., Y. Noh, and J. Dudhia, 2006: A new vertical diffusion package with an explicit treatment of entrainment processes. Mon. Wea. Rev., 134, 23182341, https://doi.org/10.1175/MWR3199.1.

    • Search Google Scholar
    • Export Citation
  • Iacono, M. J., J. S. Delamere, E. J. Mlawer, M. W. Shephard, S. A. Clough, and W. D. Collins, 2008: Radiative forcing by long‐lived greenhouse gases: Calculations with the AER radiative transfer models. J. Geophys. Res., 113, D13103, https://doi.org/10.1029/2008JD009944.

    • Search Google Scholar
    • Export Citation
  • James, E. P., and Coauthors, 2022: The High-Resolution Rapid Refresh (HRRR): An hourly updating convection-allowing forecast model. Part II: Forecast performance. Wea. Forecasting, 37, 13971417, https://doi.org/10.1175/WAF-D-21-0130.1.

    • Search Google Scholar
    • Export Citation
  • Jiménez, P. A., J. Dudhia, J. F. González-Rouco, J. Navarro, J. P. Montávez, and E. García-Bustamante, 2012: A revised scheme for the WRF surface layer formulation. Mon. Wea. Rev., 140, 898918, https://doi.org/10.1175/MWR-D-11-00056.1.

    • Search Google Scholar
    • Export Citation
  • Gallus, W. A., Jr., and M. Segal, 2001: Impact of improved initialization of mesoscale features on convective system rainfall in 10-km eta simulations. Wea. Forecasting, 16, 680696, https://doi.org/10.1175/1520-0434(2001)016<0680:IOIIOM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Koračin, D., and C. E. Dorman, 2001: Marine atmospheric boundary layer divergence and clouds along California in June 1996. Mon. Wea. Rev., 129, 20402056, https://doi.org/10.1175/1520-0493(2001)129<2040:MABLDA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Krishnamurthy, R., and L. Sheridan, 2020a: California–Wind Sentinel (120), Humboldt/reviewed data. U.S. Department of Energy, accessed 1 May 2023, https://doi.org/10.21947/1783807.

  • Krishnamurthy, R., and L. Sheridan, 2020b: California–Wind Sentinel (130), Morro Bay/reviewed data. U.S. Department of Energy, accessed 1 May 2023, https://doi.org/10.21947/1959715.

  • Krishnamurthy, R., and L. Sheridan, 2021a: California–Leosphere Windcube 866 (120), Humboldt/reviewed data. U.S. Department of Energy, accessed 1 May 2023, https://doi.org/10.21947/1783809.

  • Krishnamurthy, R., and L. Sheridan, 2021b: California–Leosphere Windcube 866 (130), Morro Bay/reviewed data. U.S. Department of Energy, accessed 1 May 2023, https://doi.org/10.21947/1959721.

  • Krishnamurthy, R., and Coauthors, 2023: Year-long buoy-based observations of the air–sea transition zone off the U.S. West Coast. Earth Syst. Sci. Data, 15, 56675699, https://doi.org/10.5194/essd-15-5667-2023.

    • Search Google Scholar
    • Export Citation
  • Lin, W., M. Zhang, and N. G. Loeb, 2009: Seasonal variation of the physical properties of marine boundary layer clouds off the California coast. J. Climate, 22, 26242638, https://doi.org/10.1175/2008jcli2478.1.

    • Search Google Scholar
    • Export Citation
  • Liu, Y., Y. Qian, and L. K. Berg, 2022: Local-thermal-gradient and large-scale-circulation impacts on turbine-height wind speed forecasting over the Columbia River basin. Wind Energy Sci., 7, 3751, https://doi.org/10.5194/wes-7-37-2022.

    • Search Google Scholar
    • Export Citation
  • Monteiro, I. T., A. J. Santos, M. Belo‐Pereira, and P. B. Oliveira, 2016: Adjustment of the summertime marine atmospheric boundary layer to the western Iberia coastal morphology. J. Geophys. Res. Atmos., 121, 38753893, https://doi.org/10.1002/2016JD025055.

    • Search Google Scholar
    • Export Citation
  • Musial, W., P. Spitsen, P. Duffy, P. Beiter, M. Marquis, R. Hammond, and M. Shields, 2022: Offshore wind market report: 2022 edition. NREL Rep., 126 pp., https://www.energy.gov/sites/default/files/2022-08/offshore_wind_market_report_2022.pdf.

  • Nakanishi, M., and H. Niino, 2009: Development of an improved turbulence closure model for the atmospheric boundary layer. J. Meteor. Soc. Japan, 87, 895912, https://doi.org/10.2151/jmsj.87.895.

    • Search Google Scholar
    • Export Citation
  • National Renewable Energy Laboratory, 2023: 2023 National Offshore Wind data set (NOW-23). NREL, accessed 23 October 2023, https://doi.org/10.25984/1821404.

  • Newsom, R., and R. K. Krishnamurthy, 2022: Doppler lidar (DL) instrument handbook. Rep. DOE/SC-ARM-TR-101, 59 pp., https://doi.org/10.2172/1034640.

  • Olson, J. B., and Coauthors, 2019a: Improving wind energy forecasting through numerical weather prediction model development. Bull. Amer. Meteor. Soc., 100, 22012220, https://doi.org/10.1175/BAMS-D-18-0040.1.

    • Search Google Scholar
    • Export Citation
  • Olson, J. B., J. S. Kenyon, W. Angevine, J. M. Brown, M. Pagowski, and K. Sušelj, 2019b: A description of the MYNN-EDMF scheme and the coupling to other components in WRF–ARW. NOAA Tech. Memo. OAR GSD-61, 42 pp., https://repository.library.noaa.gov/view/noaa/19837.

  • Optis, M., O. Rybchuk, N. Bodini, M. Rossol, and W. Musial, 2020: Offshore wind resource assessment for the California Pacific outer continental shelf (2020). NREL Rep. NREL/TP-5000-77642, 61 pp., https://doi.org/10.2172/1677466.

  • Parish, T. R., D. A. Rahn, and D. C. Leon, 2016: Aircraft measurements and numerical simulations of an expansion fan off the California coast. J. Appl. Meteor. Climatol., 55, 20532062, https://doi.org/10.1175/JAMC-D-16-0101.1.

    • Search Google Scholar
    • Export Citation
  • Payne, A. E., and Coauthors, 2020: Responses and impacts of atmospheric rivers to climate change. Nat. Rev. Earth Environ., 1, 143157, https://doi.org/10.1038/s43017-020-0030-5.

    • Search Google Scholar
    • Export Citation
  • Pronk, V., N. Bodini, M. Optis, J. K. Lundquist, P. Moriarty, C. Draxl, A. Purkayastha, and E. Young, 2022: Can reanalysis products outperform mesoscale numerical weather prediction models in modeling the wind resource in simple terrain? Wind Energy Sci., 7, 487504, https://doi.org/10.5194/wes-7-487-2022.

    • Search Google Scholar
    • Export Citation
  • Rahn, D. A., T. R. Parish, and D. Leon, 2014: Coastal jet adjustment near Point Conception, California, with opposing wind in the bight. Mon. Wea. Rev., 142, 13441360, https://doi.org/10.1175/MWR-D-13-00177.1.

    • Search Google Scholar
    • Export Citation
  • Saha, S., and Coauthors, 2010: The NCEP Climate Forecast System Reanalysis. Bull. Amer. Meteor. Soc., 91, 10151058, https://doi.org/10.1175/2010BAMS3001.1.

    • Search Google Scholar
    • Export Citation
  • Schroeder, I. D., B. A. Black, W. J. Sydeman, S. J. Bograd, E. L. Hazen, J. A. Santora, and B. K. Wells, 2013: The North Pacific High and wintertime pre‐conditioning of California current productivity. Geophys. Res. Lett., 40, 541546, https://doi.org/10.1002/grl.50100.

    • Search Google Scholar
    • Export Citation
  • Skamarock, W. C., and Coauthors, 2019: A description of the Advanced Research WRF Model version 4. NCAR Tech. Note NCAR/TN-556+STR, 145 pp., https://doi.org/10.5065/1dfh-6p97.

  • Smith, C., B. Hatchett, and M. Kaplan, 2018: Characteristics of sundowner winds near Santa Barbara, CA, from a dynamically downscaled climatology: Environment and effects aloft and offshore. J. Geophys. Res. Atmos., 123, 13 09213 110, https://doi.org/10.1029/2018JD029065.

    • Search Google Scholar
    • Export Citation
  • Soares, P. M. M., P. M. A. Miranda, A. P. Siebesma, and J. Teixeira, 2004: An eddy‐diffusivity/mass‐flux parametrization for dry and shallow cumulus convection. Quart. J. Roy. Meteor. Soc., 130, 33653383, https://doi.org/10.1256/qj.03.223.

    • Search Google Scholar
    • Export Citation
  • Zemba, J., and C. A. Friehe, 1987: The marine atmospheric boundary layer jet in the Coastal Ocean Dynamics Experiment. J. Geophys. Res., 92, 14891496, https://doi.org/10.1029/JC092iC02p01489.

    • Search Google Scholar
    • Export Citation
  • Fig. 1.

    Simulation domain for the short-term simulations. The red dots show locations of the Humboldt and Morro Bay lidar buoys. The gray dots and orange squares show the locations of the NOAA surface buoys and onshore radar used to validate the CA20 dataset. The inserts show the locations of BOEM wind energy lease areas. The black line indicates the Humboldt–McKinleyville cross section.

  • Fig. 2.

    The LLJ and drivers over the California outer continental shelf.

  • Fig. 3.

    Mean wind speed simulated using the (a),(d) YSU and (b),(e) MYNN PBL schemes and (c),(f) the difference between the MYNN and YSU simulations during July 2021. Shown are (top) 100- and (bottom) 4-m wind speeds. The circles in (a), (b), (d), and (e) are colored with observed wind speed. The red dots in (c) and (f) indicate the locations of the Humboldt (north) and Morro Bay (south) lidar buoys. Note that the reference vectors are different in (c) and (f).

  • Fig. 4.

    Observed 100-m wind speed at Humboldt (a) bias and (b) RMSE as a function of month, and (c) bias/difference as a function of wind direction using the MYNN and YSU PBL schemes. Lines are for observed 100-m wind speed.

  • Fig. 5.

    Time series of observed and simulated 100-m wind speed for (a) 12–15 Jul 2021, (b) 21–24 Jul 2021, (c) 18–21 Jul 2021, (d) 5–8 Jul 2021, and (e) 7–10 Nov 2020.

  • Fig. 6.

    As in Fig. 3, but for 1800 UTC 3 Jul 2021 and only at 100 m.

  • Fig. 7.

    Time–height cross sections of wind speed at Humboldt from the (a) lidar, (b) YSU simulation, (c) MYNN simulation, and (d) MYNN minus YSU during 12–15 Jul 2021 (note the different extent of the vertical axis). The blue curve in (a) indicates the temperature difference between McKinleyville and Humboldt at 300 m.

  • Fig. 8.

    Vertical profiles of wind speed at Humboldt at 1200 UTC (solid line) and 0000 UTC (dashed line) for (a) 12–15 Jul 2021, (b) 21–24 Jul 2021, (c) 18–21 Jul 2021, (d) 5–8 Jul 2021, and (e) 7–10 Nov 2020.

  • Fig. 9.

    Humboldt–McKinleyville cross section (shown in Fig. 1) of (a) temperature (°C) and zonal and vertical wind vector (m s−1) from YSU simulation, and (b) temperature difference between MYNN and YSU at 1800 UTC 13 Jul 2021 during case 1. (c),(d) As in (a) and (b), but for V wind (m s−1). The red dots show the location of the radar wind profiler near the coast and the offshore buoy at Humboldt. The vertical wind speed is scaled by multiplying 100.

  • Fig. 10.

    As in Fig. 9, but for 1800 UTC 19 Jul 2021.

  • Fig. 11.

    As in Fig. 9, but for 1800 UTC 6 Jul 2021.

  • Fig. 12.

    As in Fig. 9, but for 1800 UTC 7 Nov 2020.

  • Fig. 13.

    Time series of virtual temperature at McKinleyville at 300 m in July. The shaded regions show different time periods corresponding to the various case studies analyzed.

  • Fig. 14.

    Drivers associated with the LLJ in the (a) YSU and (b) MYNN simulations. The differences between the two simulations are marked in red in (b).

  • Fig. 15.

    Differences in (a) cloud fraction, (b) downward shortwave, and (c) 2-m temperature using simulations with MYNN minus YSU, averaged over July 2021.

All Time Past Year Past 30 Days
Abstract Views 2570 1794 0
Full Text Views 1338 1075 873
PDF Downloads 446 173 20