Re: [ferret_users] Along-shore and cross-shore windstress calculation

[vivek shilimkar <vivek.shilimkar@xxxxxxxxx>]

Dear All,

Now I have another concern related to the calculation of along-shore and offshore components of wind.

I calculated the direction of wind (angle) by following method.

Let rad = 4.*atan(1.0)/180

Let wdir = atan2(u,v)/rad +180

This gives the angle of wind in terms of meteorological direction.

So, easterly winds will have angle 90 degrees, westerly - 270 degrees, northerly - 360, southerly - 180 degrees.

However, when winds direction changes from northerly to north-easterly (or vice-versa), the change in direction of wind (angle) is very large (from 360 to 30 degrees)

Please check the image attached with email and observe the change in angle of wind in Arabian Sea (Indian Ocean), Atlantic Ocean (west African coast), Eastern Pacific Ocean.

​

In above three regions you can observe the above situation when northerly winds become north-easterly. 

(Color plot is wind direction (angle) and vectors are wind stress)

Should I go ahead with the above results and calculate the along-shore and offshore components?

Or is there any way to treat such change in direction in better way?

Regards!

Vivek

On Thu, Jun 28, 2018 at 5:56 PM vivek shilimkar <vivek.shilimkar@xxxxxxxxx> wrote:

Dear Dr. Ryo and William,

Thank you so much for your reply.

I understand the Dr. William's method very quickly and prefer to use that.

Regards!

Vivek

On Wed, Jun 27, 2018 at 11:40 PM Ryo Furue <furue@xxxxxxxxxx> wrote:

Vivek and Billy,

On Wed, Jun 27, 2018 at 9:31 PM William Kessler <william.s.kessler@xxxxxxxx> wrote:

Attached is a scan of some old notes, not code. It's pretty straightforward to code once you do the trigonometry (that's what's in these notes).

My method (https://www.pmel.noaa.gov/maillists/tmap/ferret_users/fu_2013/msg00345.html) just calculates cos θ and sin θ in Billy's notes.  My idea is based on the fact that the "coastline" is the isobath of z = 0.  So, if you have an elevation/bathymetry data like ETOPO5, you can calculate the direction (θ) of the isobath from the gradient of the elevation.

All Billy's caveats apply to my method, too.

 

The trickiest part of a calculation like this is deciding what you mean by "coast": Is it the wind at the beach? Or 25, 50, 100km offshore? 

That has implications for how complicated your "coastline" is, but more importantly the wind is almost always stronger further offshore, so the result depends very much on this choice. This is further complicated by how the wind product was calculated and how that processing treated the nearshore region.

The scale of coast you choose will therefore have large implications for the result. 

I guess the first question you want to ask is: Is a straight line good enough as an approximation to your coastline? If your answer is yes, then determine a straight line that best approximates your coastline and then you know the angle θ, which is a single number for the entire calculation.

Ryo

Attachment: wind-direction_vector.gif
Description: GIF image