Status report for PACJET-2002 field activities

23 January 2002
Marty Ralph
NOAA/Environmental Technology Laboratory

See PACJET 2003 for information about the 2003 field season.

See PACJET 2001 for information about the 2001 field season.

A Brief Background

  • PACJET-2002 is the third in a series of field experiments, preceded by CALJET in 1997/98 and PACJET in 2001, to focus on improving short-term forecasts of land-falling Pacific winter storms on the U.S. West Coast.
  • As with earlier years, the effort focuses on both research and direct operational applications.
  • Results from CALJET and PACJET-2001 are appearing in conferences, journals, and patents, and have identified issues that help focus the 2002 field phase.
  • Input at the PACJET-2001 Workshop in Monterey from 24-26 October contributed to refining both short-term goals for winter 2001/2002 and for long-term program development.
  • PACJET-2002 incorporates ground-based, satellite, aircraft and modeling activities.
  • Data collected in PACJET-2002, which is characterized as a neutral phase of ENSO, will complement those gathered in the strong El Niño of 1997/98 and the weak La Niña of 2001.

Research Priorities in 2002 Data Collection

  • Assessing QuickScat satellite winds in the low-level jet: P-3 aircraft activities will focus on evaluation of QuickScat satellite-derived surface winds and on exploring how conditions aloft in the boundary layer may be inferred. This will increase the sample-size of high wind conditions for assessing model PBL parameterizations. This effort is partly funded by NASA (Persson and Walter) and is coordinated with NOAA's Ocean Winds Program (Chang).
  • Assessing the value of GOES rapid scan winds in mesoscale data assimilation: The RUC model (Benjamin and Weygandt) is being run at 10-km resolution in the west, and will be used retrospectively to assess the impact of assimilating rapid scan GOES winds as part of the GOES Winds Experiment-GWINDEX (Velden, Stettner, Daniels, Bresky).
  • Documenting the frequency of occurrence of non-bright band rain found in CALJET: Results from a new ETL high-dynamic range S-band radar deployed in CALJET indicate that roughly 30% of all rain that fell at a key coastal site that year occurred without a bright band (White et al. 2002), however <5% did so in 2001. The same radar has returned to the site in 2002 and another is being used in the western Sierra Foothills to extend this study.
  • Exploring the Weather-Climate Connection: The severity of flooding in specific watersheds is very sensitive to low-level wind conditions, which vary with the phase of ENSO (Ralph et al. 2002). Coastal wind profilers are being used to add to data collected in other phases of ENSO during CALJET and PACJET-2001. These will provide ground truth for assessing the accuracy of NCEP reanalyses in this data sparse area for this application.

Forecast Applications Priorities in the 2002 Field Phase

  • Direct forecaster use of experimental data will be systematically evaluated (Nance): Forecaster use of wind profiler, GOES rapid-scan winds, and the RUC model will be tracked through an on-line feedback form, referencing in area forecast discussions, and by other means. Plans for training are being developed. Due to funding limitations for staff needed on the P-3 and on the ground, P-3 data will be available in real time on a more limited basis.
  • GOES Rapid-Scan Winds-the GWINDEX Experiment (Velden): In addition to the research associated with documenting the impact of assimilating GOES rapid scan winds in the RUC model as described above, the data are viewable on line for direct forecaster use, e.g., evaluating the position and intensity of key wind features using the enhanced capabilities of the rapid scan winds.
  • The Rapid Update Cycle model (RUC) (Benjamin and Weygandt): A special version of the RUC model is being run that assimilates the GWINDEX data every 3 hours, and includes a nest in the west with 10-km grid spacing. These data will be available directly on the standard AWIPS workstations that are in use in NWS Weather Forecast Offices.
  • Experimental wind profiler array (Ralph, Neiman and White): Research results from CALJET and PACJET have quantified the importance of the low-level jet in determining rain rates in coastal storms, and initial experience with the use of a prototype real-time bright-band snow-level detection algorithm developed for PACJET-2001 will be extended to additional sites. Eleven profilers will cover the coast from S. California to Washington State.

Overview of major components of field deployments in PACJET 2002

Ground-based, satellite, modeling
Dates for RUC, profilers, GWINDEX:14 January - 31 March 2002
S-band vertically pointing radars:Cazadero and Grass Valley, CA.
ETL and NPS Wind profiler sites:
  • Newport, OR
  • Westport, WA
  • Goleta, CA
  • Goleta, CA
  • Piedras Blancas, CA
  • Monterey (NPS), CA
  • Bodega Bay, CA
  • Eureka, CA
  • Grass Valley, CA
  • Chico, CA
  • Waterford, CA
  • Lost Hills, CA
Dates for P-3 deployment:1 February - 2 March 2002
Flight hours for P-3:100 h (70 h PACJET + 30 h Ocean Winds)
Location of aircraft operations base:Portland, OR
P-3 operations area:Offshore from California to Vancouver, Canada

Detailed descriptions of key activities

  • Ground-based observations (Ralph, Neiman, White):
    Real-time data and a site map can be seen at
  • The GOES Rapid-Scan Winds Experiment-GWINDEX (Velden, Stettner at UW-CIMSS, and Daniels and Bresky at NESDIS/FPDT):

    Data web link:
    Overall Objective: To demonstrate improved quantity and quality of cloud-motion winds using 7.5 minute rapid-scan visible and infrared imagery from GOES-10.
    Goals: To provide improved remotely-sensed data over the Eastern Paciafic for NWS forecasters, support PACJET and THORPEX initiatives, and assess data impact on the RUC model short-term forecasts.
    Experiment Duration: January 8 through March 31, 2002.
    Coverage Domain: Eastern North Pacific and west coast North America.
    Data Set Availability: Three hourly, around the clock.
    Real time: roughly 1 hour after image sequence.
    Participants: NOAA/NESDIS/ORA/FPDT and UWisc.-CIMSS, NWS, NOAA/ETL, NOAA/FSL, PACJET community.
    Point of Contact: Chris Velden (UW-CIMSS)

  • RUC modeling activities (Weygandt and Benjamin/FSL):

    In support of PACJET 2002, the Regional Analysis and Prediction Branch of FSL is running a special high-resolution version of the Rapid-Update Cycle (RUC) model and distributing forecast fields to the NWS Western Region Headquarters for dissemination to local offices. The special RUC configuration consists of a 10-km grid covering all of the NWS Western Region, nested within a 20-km CONUS grid. For the 20-km CONUS grid, 24-h forecasts are produced every 3 h using a 1-h data assimilation cycle. In addition to ingesting all conventional observations, satellite-derived cloud-motion vectors determined at 3-h intervals (provided by C. Velden, U. Wisc./CIMMS) and special mesonet (including MesoWest) observations are being assimilated into the 20-km forecast cycle. For the 10-km grid, 24-h forecasts are produced every 6 h using the 20-km CONUS RUC as lateral boundary conditions. Forecast products can be viewed on the PACJET RUC web-page.

    The goals of this effort are to:

    1. provide operational NWS forecasters with real-time guidance products from a next-generation version of the RUC model,
    2. obtain forecaster feedback on strengths and weaknesses of the model guidance products,
    3. evaluate the impact on forecast skill from ingesting data from new observing systems.
    This last goal has been achieved by conducting controlled retrospective data sensitivity experiments. Initial results from two 3-day periods during PACJET-2001 showed that inclusion of the rapid-scan satellite wind observations reduced RMS errors in short-range (0-12 h) wind predictions by up to 10%.

  • NOAA P-3 Aircraft activities (Persson and Ralph): The P-3 will be based in Portland, Oregon from Feb. 1 - March 2. The operations center will be in the Portland NWS Weather Forecast Office where weather briefings will be held at 10:30 AM PT. Contact Ola Persson for further details or updates.

Science Objectives
The scientific foci of this year's program are:

  1. Validation of the low-level jet (LLJ) structure seen by satellite (QuickScat) overpasses, and obtaining measurements of the thermodynamic and kinematic structure of the low-level jet region to be used for extrapolating the satellite measurements. An ETL/CIRES objective.
  2. Validation of NESDIS/UMass scatterometer(s) and microwave radiometer in high winds and precipitation - obtain real-time surface winds and range profiling of the effects of precipitation on the measurements. Some problems with the scatterometers upgrades will narrow some of these objectives.
  3. Obtaining Doppler radar and in-situ measurements of the NCFR to improve our understanding of the dynamics of the core and gap regions. An ETL/NSSL objective.
  4. Participate in one objective targeting IOP (super-IOP) in collaboration with NCEP-this objective still requires support for the dropsonde costs.
  5. Obtaining GPS backscatter surface winds - NASA project
All science objectives, with the exception of 4), can be met by flying in the vicinity of the cold front and the low-level jet. The first objective requires in-situ sampling of the LLJ region at ±2 hours from satellite overpass times. These overpass times occur at about 0630 and 1830 local time (local to the overpass region). Typically, this means that we will want to be sampling at low levels in the LLJ between 0430-0830 or 1630-2030 local time. With an estimate of a 1-hour ferry time, and 1-2 hours of dropsonde measurements for science objective 2) before descending to low levels, typical takeoff times will be between 01-03 and 1300-1500 local Portland time (09-11 and 21-23 UTC). The key logistical forecasting problem for each day will be to determine where the prefrontal low-level jet will be relative to the satellite overpass swath at the time of the overpass, and how strong will this jet be.