Marine Environment Radio-Sonde Verification of High Resolution Mesoscale MM5 Model Runs   OC 3570 Project By LCDR Jimmy Horne

Importance Most modern forecast use model analysis to make predictions of the future atmospheric conditions. All too often, model output is taken without considering the physical discrepancies inherent to the physics or parameterizations used. Even when considering forcing characteristics in the horizontal, the vertical can prove to be significant. -See above; The sea breeze also influences coastal currents, which affect coastal erosion and beach development, coastal marine ecosystems, and activities such as fishing and shellfishing. It is one of the most studied atmospheric phenomena due to the heavy concentration of the world’s population along the coastline of oceans (Banta et al, 1993). I choose to study the sea breeze because it is a part of my thesis work, which I begin next quarter with Dr. Nuss. The purpose of this project was to collect, compare, and analyze data from the summer cruise.

Background Originally research project was based on remote sensing applications of the marine boundary layer and their verification with actual observed conditions. Overcast conditions limited the useful data for that topic so plan ‘B’ went into action. The basic forcing mechanism driving sea breeze circulations is the development of a coastal thermal gradient to which the wind responds. A horizontal temperature gradient develops due to the differential radiative properties of the water and land surfaces. This temperature gradient will produce a pressure gradient. The warm air over land tends to lower the surface pressure relative to the unchanged or cooler boundary layer air over the water. Thus, an onshore directed pressure gradient arises to which the air must respond by accelerating toward the coastline Sea breezes typically form and begin their flow inland during the mid- to late-morning hours, when daytime temperature gradient between land and ocean is fairly strong. The sea breeze circulation intensifies as solar heating reaches its maximum. Afternoon is the most active time of day for this circulation, as sea breeze penetration reaches a maximum and winds are strongest. An abundance of radio-sonde data combined with a persistent inversion provided an ideal opportunity to study model features in a synoptically stable environment.

Background MM5 numerical model runs hosted by the Naval Postgraduate School Department of Meteorology were use to compare with radio-sonde data collected during the second leg of the OC3570 cruise. This endeavor allowed a realistic comparison of typical model field to several up/down sonde launches. Revealed some interesting aspects of fields within the marine boundary layer. The basic forcing mechanism driving sea breeze circulations is the development of a coastal thermal gradient to which the wind responds. A horizontal temperature gradient develops due to the differential radiative properties of the water and land surfaces. This temperature gradient will produce a pressure gradient. The warm air over land tends to lower the surface pressure relative to the unchanged or cooler boundary layer air over the water. Thus, an onshore directed pressure gradient arises to which the air must respond by accelerating toward the coastline Sea breezes typically form and begin their flow inland during the mid- to late-morning hours, when daytime temperature gradient between land and ocean is fairly strong. The sea breeze circulation intensifies as solar heating reaches its maximum. Afternoon is the most active time of day for this circulation, as sea breeze penetration reaches a maximum and winds are strongest.

Developed at Penn State and NCAR Background MM5 Model: Developed at Penn State and NCAR Popular research tool used at several universities and government laboratories A limited-area, non-hydrostatic, terrain-following sigma-coordinate model Designed to simulate or predict mesoscale and regional-scale atmospheric circulation The basic forcing mechanism driving sea breeze circulations is the development of a coastal thermal gradient to which the wind responds. A horizontal temperature gradient develops due to the differential radiative properties of the water and land surfaces. This temperature gradient will produce a pressure gradient. The warm air over land tends to lower the surface pressure relative to the unchanged or cooler boundary layer air over the water. Thus, an onshore directed pressure gradient arises to which the air must respond by accelerating toward the coastline Sea breezes typically form and begin their flow inland during the mid- to late-morning hours, when daytime temperature gradient between land and ocean is fairly strong. The sea breeze circulation intensifies as solar heating reaches its maximum. Afternoon is the most active time of day for this circulation, as sea breeze penetration reaches a maximum and winds are strongest.

MM5 Model: High resolution 12 km grid Background MM5 Model: High resolution 12 km grid The basic forcing mechanism driving sea breeze circulations is the development of a coastal thermal gradient to which the wind responds. A horizontal temperature gradient develops due to the differential radiative properties of the water and land surfaces. This temperature gradient will produce a pressure gradient. The warm air over land tends to lower the surface pressure relative to the unchanged or cooler boundary layer air over the water. Thus, an onshore directed pressure gradient arises to which the air must respond by accelerating toward the coastline Sea breezes typically form and begin their flow inland during the mid- to late-morning hours, when daytime temperature gradient between land and ocean is fairly strong. The sea breeze circulation intensifies as solar heating reaches its maximum. Afternoon is the most active time of day for this circulation, as sea breeze penetration reaches a maximum and winds are strongest.

Method of Investigation: Extract data from radio sonde lauches during leg two of the cruise Using Lat/Lon positions, extract sythetic sonde data from model run at closest grid point Compare data between the measurements and model data (forecast and analysis) for the closest launch time Analyze temperature and dew point vs. height, as well as wind speed and direction Those measurements included: 1) rawinsonde profiles 2) ship’s sail data 3) surface stations

Sonde Positions 17-24 16 Surface-station measurements include: the NPS 915 MHz Profiler located at Fort Ord, station MRY (Monterey) located at 36.59o -121.84o, and station SNS (Salinas) located at 36.66o –121.61o. Rawinsonde launches were taken at the above locations. It would have been nice to access data from a more inland and southern surface station, however it could not be acquired from Dr. Nuss. 15

Sonde Positions 16 17-24 Surface-station measurements include: the NPS 915 MHz Profiler located at Fort Ord, station MRY (Monterey) located at 36.59o -121.84o, and station SNS (Salinas) located at 36.66o –121.61o. Rawinsonde launches were taken at the above locations. It would have been nice to access data from a more inland and southern surface station, however it could not be acquired from Dr. Nuss. 15

Data Collection Sonde Model wind speed (m/s) direction (deg) temperature (C) dew point (C) relative humidity (%) pressure (hPa) ascent rate (m/s) height above MSL (m) RI MRI vapor pressure Model u (m/s) v (m/s) w (m/s) T (K) P (mb) Z (m) RH (%)

Important Note Since MM5 is a regional model, it does require initial conditions as well as lateral boundary conditions therefore must be coupled with global models and other regional models to use their output either as first guess for objective analysis, or as lateral boundary conditions. To produce lateral boundary condition for a model run, gridded data is needed to cover the entire time period that the model is integrated. All runs were initialized off 1 Degree AVN warm starts.

Analysis Sonde 15 27 hr forecast Temperature too high at surface Inversion defined too low

Analysis Sonde 15 27 hr forecast Temperature too high at surface Inversion defined too low

Analysis Sonde 15 27 hr forecast Direction 90 deg out above 600 m Speed tracking well but too high at surface

Analysis Sonde 15 15 hr forecast Temperature too high at surface Not defining structure well Too much moisture aloft Inversion defined too low

Analysis Sonde 15 15 hr forecast Direction 90 deg out above 600 m, low speeds account for poor tracking below 500 m Speed tracking well

Analysis Sonde 15 3 hr forecast Better structure A little too much moisture aloft

Analysis Sonde 15 3 hr forecast Direction 90 deg out above 700 m, speeds make coarseness below 500 m acceptable Speed tracking well with exception at 700 m

Analysis Model Comparison Overall improvement in model structure

Analysis Sonde 15 27 hr forecast Direction 90 deg error above 700 is a persistent Speed tracking well but one would expect increased accuracy with increased wind speeds

Special Case Sonde 16

No inversion in structure Analysis Special Case No inversion in structure Notice level at which model begins (≈ 250 m)

Analysis Special Case Low levels not resolved well in speed or direction

Analysis Special Case Closest Forecast (6 hour) Structure still not resolved well Too warm at lowest level

Analysis Special Case 6 hr forecast Never got a handle on speeds or direction

Analysis Off Moss Landing Example of Sonde 18, one of six in same area

Analysis Off Moss Landing Model comparison

Reason for problems with this run: Closest grid point to lat/lon of sonde launch was over land and compounded by the proximity of coastal mountain range.

Conclusions Trends observed in analysis were fairly consistent for 7 of 8 sondes. Forecast beyond 12 hours typically represented the marine inversion being much lower than actual observations Temperatures where typically too high at the surface ( ≈ 2˚ C) below the inversion and became too cool above the inversion ( ≈ 2˚ C) Wind direction showed a trend for being 90˚ out from observed winds. Since the synoptic picture doesn’t seem to indicate any reason for this occurring, it does imply that a mesoscale influence that the model was unable to resolve were most likely the cause of this discrepancy. Wind speeds trended well for the most part but it was noted on several occasions that surface speeds where 2-4 m/s faster than those observed.

Conclusions Model runs for 6 hour forecast, and less, consistently showed better inversion structure while resolving inversion heights more accurately. Wind speed and wind direction problems addressed above persisted in these forecast and analysis as well.

Conclusions MM5 has been documented to have problems in resolving boundary layer issues, especially over a marine layer or complex terrain, however it continues to be used operationally despite these failings. It is often assumed that higher resolution will yield better results. This project clearly shows that with out the correct physics, that is not necessarily true. Persistent errors with the wind speed and direction indicate the model has problems with resolving mesoscale feature in the coastal environment. YEAH! IT’S DONE!