June 07 Model-Data Comparison

Table of Contents:

  1. Initial Conditions
  2. Comparisons
    1. Plan View
    2. Point-by-Point
      1. Height
      2. Speed
        1. Speed from aircraft profiles and sondes
        2. SSM/I wind speeds
  3. Other data sets to compare the model output to in the future
  4. Some ideas for using the model output to understand the MABL

1)    Initial conditions

2)    Comparisons

2a)    Plan View Comparisons

The following plots show data and model outputs on the same domain and with the same color scale for ease of comparison.   I've masked out the interpolated observation fields far from the observation locations.  The observations include those used to initialize the model along a line upstream from Cape Mendocino.  The model data was interpolated onto a regular grid, which may smear out sharp gradients such as those across the shock.

2b)    Point-by-Point Comparisons

1)    MABL height

A comparison of all model and observed MABL heights can be observed here.   It shows that the magnitudes are realistic though scattered around a 1:1 line.

A comparison along individual sections is below.
The map shows 9 sections along which the model and observed MABL heights were compared.  The comparison plots are below the map.  Where the observations are dense, the MABL height comes from the lidar data and was extracted by Linda Strom.  Isolated observations are from aircraft profiles.


 

2)    Speed

2a)    Speed from aircraft profiles and sondes

A comparison of all model and observed speeds (from profiles and the few sondes which were functional) can be accessed here.

A comparison along individual sections is below.
The map shows 6 sections along which the model and observed speed were compared.  The comparison plots are below the map.


 

2b)    Speeds from SSM/I satellites

In the table below are comparisons between the model layer-averaged speeds and surface speeds from the SSM/I satellite for morning and evening passes.  A rough conversion is that surface winds = 75% layer averaged winds.
 
Morning pass Evening pass
Section 1 Section 1
Section 2 Section 2
Section 3 Section 3
Section 4 Section 4
Section 5 Section 5
Section 6 Section 6
Section 7 Section 7

3)    Other data to compare to the model outputs in future

4)    Using the model output to understand the MABL

Some things we could learn from the MABL using the model output combined with observations:
  1. Fluxes:
    1. Can we get a good estimate of fluxes on a specific day by using bulk formulae for fluxes + model wind field + remotely observed SST?
    2. How does this estimate compare to fluxes directly estimated from the 30 m aircraft runs?
    3. Does this estimate capture any of the complicated horizontal structure observed in the fluxes, which is not captured by estimates using bulk formulae and buoy data?
  2. Extent of the supercritical region:
    1. Can we understand the large ~500 km wide region of fast winds seen in the Nelson (1977) climatology as a supercritical or nearly critical region?  COAMPS summertime average of Froude number shows a Fr > .8 region which is ~500 km wide.
    2. Can we show by scaling or momentum balances that the supercritical/ transcritical region can extend far beyond the Rossby radius?   That the gradient balance holds beyond the Rossby radius?   Can scaling give us the length scale Lsup = 500 km?
    3. Is this true for climatological values of forcing?  Within which range of forcing?
    4. What limits the extent of the supercritical region, if not the Rossby radius: our background conditions (deeper/slower offshore)?  Decreasing centrifugal acceleration and local enhancement to the pressure gradient with distance from shore, while friction and the large scale pressure gradient are unchanged?
  3. The sub to supercritical transition:
    1. Burk et al (in press) calculated the momentum balance in the flow around Cape Mendocino, showing that a gradient balance holds.  Both sub and supercritical flow to thin/accelerate around the bend.
    2. Can we estimate the momentum balance from the model and observations to show that this is why the flow becomes supercritical?
  4. Bend flow dynamics
    1. Do the observations show secondary flow in the bend (Geyer, 1993) which leads to enhanced vertical mixing (Siem and Gregg, 1997), since the flow is in gradient balance?  The secondary flow (flow is towards the bend at the bottom and away from the bend near the top) occurs because friction reduces the speed near the bottom, while the pressure gradient remains the same; the opposite flow occurs at the top to compensate.   The model is layer averaged, so it can't be used to look at the secondary flow.
    2. Do the conditions on any flight days, such as 6/12, meet the criteria for flow separation and eddy generation around a headland? (Signell and Geyer, 1991)
    3. Can the analytical solution of Cherniawsky and Leblond (1986), which shows sub- to supercritical transition around a bend, be applied to our problem?  The paper is pretty tough (several peturbation solutions), but applicable.  Can we construct a potential flow/boundary layer solution as in Signell and Geyer (1991) to show acceleration around the cape?

 

The data on this page is unpublished. If it is used please cite the author Kathleen Edwards, the Center for Coastal Studies, and the Coastal Waves group at Scripps Insitution of Oceanography.

Please send comments or questions to me at kate@coast.ucsd.edu