[Met_help] adding a new data source

Tue Jul 14 14:29:15 MDT 2009

Joe,

I took a look at the sample file you sent, but I'm pretty confused about
the data.

In the header for that file, it seems to indicate that the wind speeds are
measured in a particular direction.  For example, the wind speeds at 30m,
50m, and 60m are relative to 270 degrees.  At 59m, wind speeds are
relative to 240 degrees.  And at 2m and 3m, they're relative to 360
degrees.  Am I understanding that correctly?

When verifying wind speed, MET currently looks in the forecast file for a
wind speed record.  If one exists, it uses it.  If not, it grabs U and V,
rotates from grid-relative to earth-relative, if necessary, and computes
the wind speed field from U and V.  But are we able to compare these wind
speeds to the observations that are defined "relative" to a particular
direction?

Do you see a problem here?

For the vertical levels relative to 360, do you actually want to be
comparing to the V-component of wind?  And for those relative to 270, do
you want to be comparing to the U-component?

And just to clarify...
Are the "Anemometer" speeds in m/s?
Are the "Wind Vane" values directions where 0 degrees = north?
What is a "Pyranometer" and what do the values mean?

And do you want to just use the "avg" value as the observation or would
you prefer to take into account the other measures like min, max, std, and
dq?

Thanks,
John

> Thanks John,
>
> Sorry about that. That legal agreement ONLY took us 8 months to work
> out! Attached is a sample file which has data for the entire month of
> Aug 2007 for a given station.  Not all files they gave me are
> identical format, but this one seems representative, having each
> variable in the form of an average, standard deviation, max, min, and
> dq(?). Each line has the following format:
>
> STID,Date,Time,60m_Anemometer_avg,60m_Anemometer_std,
> 60m_Anemometer_min,60m_Anemometer_max,60m_Anemometer_dq,
> 59m_Anemometer_avg,59m_Anemometer_std,59m_Anemometer_min,
> 59m_Anemometer_max,59m_Anemometer_dq,50m_Anemometer_avg,
> 50m_Anemometer_std,50m_Anemometer_min,50m_Anemometer_max,
> 50m_Anemometer_dq,30m_Anemometer_avg,30m_Anemometer_std,
> 30m_Anemometer_min,30m_Anemometer_max,30m_Anemometer_dq, 0m_No
> Sensor_avg, 0m_No Sensor_std, 0m_No Sensor_min, 0m_No Sensor_max,
> 0m_No Sensor_dq, 0m_No Sensor_avg, 0m_No Sensor_std, 0m_No Sensor_min,
> 0m_No Sensor_max, 0m_No Sensor_dq,60m_Wind Vane_avg,60m_Wind Vane_std,
> 60m_Wind Vane_min,60m_Wind Vane_max,60m_Wind Vane_dq,30m_Wind Vane_avg,
> 30m_Wind Vane_std,30m_Wind Vane_min,30m_Wind Vane_max,30m_Wind
> Vane_dq, 2m_Temp_avg, 2m_Temp_std, 2m_Temp_min, 2m_Temp_max,
> 2m_Temp_dq, 3m_Pyranometer_avg, 3m_Pyranometer_std,
> 3m_Pyranometer_min, 3m_Pyranometer_max, 3m_Pyranometer_dq, 0m_No
> Sensor_avg, 0m_No Sensor_std, 0m_No Sensor_min, 0m_No Sensor_max,
> 0m_No Sensor_dq, 0m_No Sensor_avg, 0m_No Sensor_std, 0m_No Sensor_min,
> 0m_No Sensor_max, 0m_No Sensor_dq
>
> -joe
>
>
>
>
>
> On Jul 9, 2009, at 7:14 AM, John Halley Gotway wrote:
>
>> Justin and Michael,
>>
>> I read Iberdrola's NDA that Joe Olson sent me and will abide by it.
>> I understand that the use of any sample data Joe will send me falls
>> under this NDA, and I will delete that data once I've resolved
>> Joe's issues.  Also, I won't transfer or distribute this data to any
>> one else while it is in my possession.
>>
>> Thanks,
>> John Halley Gotway
>> johnhg at ucar.edu
>>
>> Joe Olson wrote:
>>> Hi John,
>>>
>>> Thanks for the quick reply and I am happy to have your help. Since
>>> the wind
>>> tower data
>>> is proprietary and required a non-disclosure agreement between NOAA
>>> and Iberdrola,
>>> I had to check with our Iberdrola contacts before sharing a sample
>>> file. They
>>> were fine
>>> with it if you provide an email stating that you understand that
>>> the use of
>>> this data falls
>>> under the same NDA as our usage and that the data needs to be
>>> deleted after use and
>>> not transferred t o any ad Iberdrola's consent. So basically, I'd be
>>> happy to send you a sample file, but you will need to browse the
>>> attached NDA and
>>> send and email to Justin & Mike (CC'd) confirming that you will not
>>> violate it.
>>> Sorry about
>>> this...
>>>
>>> As to your other questions:
>>>
>>> We want to match observations that are exactly 20, 40, or 60 m AGL.
>>>
>>> I think the wind tower data files have elevation in terms of MSL
>>> and the height
>>> of the
>>> instruments AGL, so if I understand you correctly, it does make
>>> sense to compute
>>> the
>>> AGL for the observation as "Height" minus "Elevation".
>>>
>>> Yes, I need to interpolate the model data to the observation
>>> heights. It would
>>> be nice if
>>> we had more than 5 model layers in the lowest 150 m, but since we
>>> don't, I'll
>>> have to
>>> perform some type of
>>> -joe
>>>
>>>
>>> ------------------------------------------------------------------------
>>>
>>>
>>> On Jul 8, 2009, at 8:21 AM, John Halley Gotway wrote:
>>>
>>>> Joe,
>>>>
>>>> In METv2.0, Point-Stat is not set up to perform verification at
>>>> 20, 40, and 60
>>>> m AGL.  Instead it performs verification either at the surface
>>>> (i.e. 2-m temp,
>>>> 10-m winds) or at pressure levels.  The
>>>> matching in the vertical is rather simplistic...
>>>> - Any observations with a message type of ADPSFC or SFCSHP are
>>>> considered to
>>>> be at the surface.
>>>> - When verifying upper air forecasts, we interpolate the forecast
>>>> values above
>>>> and below the observation to the observation pressure level, doing
>>>> so linear
>>>> in the log of pressure.
>>>>
>>>> We've made some minor changes in the logic to the development
>>>> version of MET
>>>> to enable him to verify forecasts of soil temperature below
>>>> ground.  An d
>>>> those k for you as well.
>>>> But I'm guessing it'd need some tweaking.
>>>>
>>>> I'd need to get more detail from you about what sort of logic
>>>> you'd need.
>>>> Here are some questions off the top of my head:
>>>> - When verifying winds at say 20 m AGL, do you want to match to
>>>> observations
>>>> that are exactly 20 m AGL, or do you need to match to a range...
>>>> like 18 to 22?
>>>> - The contents of the point observation information is described
>>>> here:
>>>> http://www.dtcenter.org/met/users/support/online_tutorial/METv2.0/ascii2nc/index.php
>>>> We store the "Elevation" of the observing location in terms of
>>>> MSL, and we
>>>> store the "Height" of the observation in terms of MSL.  Does it
>>>> make sense to
>>>> compute the AGL for the observation as
>>>> "Height" minus "Elevation"?
>>>> - Do you want be able to do any sort of vertical interpolation
>>>> down to the
>>>> observation level?  For example, if you have an obs ervation ike to
>>>> interpolate the forecasts at 20
>>>> and 40 to the observation location?  That'd be more work.
>>>>
>>>> Could you send me a sample forecast file containing forecasts at
>>>> 20, 40, and
>>>> 60 m AGL?  And could you send me some sample observation wind
>>>> tower data?  I'm
>>>> curious to see the format of it.
>>>>
>>>> Thanks,
>>>> John
>>>>
>>>>
>>>> Joe Olson wrote:
>>>>> Hi,
>>>>>
>>>>> I'd like to verify WRF simulations with some wind tower data I have
>>>>> collected. Is there any documentation I missed that may help? As a
>>>>> start, I would like to use pointstat to verify the winds &
>>>>> temperatures @ 20, 40, and 60 m AGL. I can transform the wind tower
>>>>> data to netcdf (after I learn the exact format) myself. But wh at
>>>>> I &nb
>>>>> kquote type="cite">could really use help on is which files I
>>>>> should focus on
>>>>> modifying in
>>>>> order to ingest this data.  Any pointers would be greatly
>>>>> appreciated.
>>>>>
>>>>> Thanks,
>>>>> Joe Olson
>>>>> _______________________________________________
>>>>> Met_help mailing list
>>>>> Met_help at mailman.ucar.edu <mailto:Met_help at mailman.ucar.edu>
>>>>> http://mailman.ucar.edu/mailman/listinfo/met_help
>>>
>
>