Use of Scanning Doppler Lidars during WFIP-2 to Characterize Flow Phenomena and Evaluate NWP Model Performance

As part of the Second Wind Forecast Improvement Project (WFIP-2) NOAA/ESRL has deployed two scanning, pulsed Doppler lidars to the Columbia River Basin of Oregon/Washington for the 18-month project duration from September 2015 to March 2017. Project goals include measurements to better understand and forecast wind-flow phenomena that impact wind energy, and improving numerical weather prediction (NWP) forecast models. The lidars are operated remotely 24/7/365, providing real-time profile and other wind data to a web site for immediate project use. An important aspect of this web site for the project objectives related to NWP verification and improvement, is the near real time calculation of model errors, including time series of rms model wind-speed errors for models such as NOAA’s RAP and HRRR. Scan data and calculated lidar-measured profile data, as well as the model error data, provide important information on wind phenomena encountered during the project, including frontal passages, mixout events, and mountain waves and wakes. This presentation focuses on certain summertime phenomena observed during the project in 2016: the diurnal sea breeze cycle, traveling shortwave troughs, and the marine push or surge. The lidars and other deployed instrumentation characterize the structure, evolution, and case-to-case variability of these phenomena and evaluate NWP model strengths and weaknesses. For example, the lidars often found clear density-current structure to the sea-breeze fronts, and also found that the models routinely start the sea-breeze flow too soon and end it too abruptly, leading to large errors often exceeding 5 m/s at these times. Because sea-breeze forcing, resulting from land-sea surface heating contrasts, is relatively well understood and straightforward, these cases should provide opportunities to trace the sources of many model errors.