I couldn't find the exact method of how the dropsonde winds are calculated, such as over a time period. I tried testing raw data but the ASPEN program knows when I am manipulating data and because I don't know much about the program I couldn't create scenarios where I could have faked numbers to see how it averages the data. I finally gave up trying to find an exact number of seconds. It's definitely a peak wind though. While I was skimming through stuff I found some various papers that others might find interesting. (a lot is really technical, but there is some interesting stuff in them between the technical stuff)
----
The NCAR GPS Dropwindsonde:
http://journals.ametsoc.org/doi/abs/10.1175/1520-0477%281999%29080%3C0407%3ATNGD%3E2.0.CO%3B2
----
GPS Dropwindsonde Wind Profiles in Hurricanes and Their Operational Implications:
http://journals.ametsoc.org/doi/abs/10.1175/1520-0434%282003%29018%3C0032%3AGDWPIH%3E2.0.CO%3B2
----
A Reanalysis of Hurricane Andrew's Intensity:
http://journals.ametsoc.org/doi/abs/10.1175/BAMS-85-11-1699
"While individual GPS dropwindsondes provide only a momentary slice of data (which is not even a vertical profile because of the inflowing and swirling flow that the drop encounters in the hurricane eyewall), a judicious partitioning and averaging of the dropwindsondes can provide useful wind mean wind conditions within the hurricane."
"b) the dropwindsondes do not measure a 1-min wind, but measure whatever turbulence through which they happen to fall during a 0.5-s interval. These factors contribute to the high variance in adjustment-factor ratios computed from individual dropwindsondes, but it has very little to do with the relationship between peak winds from level to level."
----
Vaisala Dropsonde RD94:
http://www.vaisala.com/Vaisala%20Documents/Brochures%20and%20Datasheets/RD94-Dropsonde-Datasheet-B210936EN-A-LoRes.pdf
"The descent speed of the RD94 is approximately 11 m/s at the sea level, and about 21 m/s at a 12 km altitude. In strong convective situations the descent rate can vary a lot, and even negative (descent) rates are measured. The RD94 measures the PTU (pressure, temperature, relative humidity) twice and wind speed and direction four times a second. This means that e.g. the last measured pressure level is at maximum about 5 - 6 meters above the surface."
----------------------------------------------------------------------------
It might not be a set time limit. There is a lot of processing that is done to the raw dropsonde data. There could be a thousand observations on the way down. Then a program processes that for errors and seems like it can fill in the blanks, perhaps with equations that aren't based on an exact timeframe. But, it is definitely a very short time frame, whether it is an exact amount of time or variable. I would guess 10 seconds or less for sure. Probably around 5 seconds. It seemed that from what little I could so with the ASPEN program to fake data, it seemed like it might be around 5 seconds but I don't know. (or its variable and it is simply not very many seconds, somewhere near 5)
----------------------------------------------------------------------------
I found one interesting unrelated tidbit from the "A Reanalysis of Hurricane Andrew's Intensity" article:
"A key issue confronting both operational estimates and postanalysis assessments of hurricane intensity is the most appropriate way to adjust flight-level winds typically at 700 mb down to surface wind values. Ten second averages of the flight-level wind in the inner core of hurricanes are assumed to represent a 1-min-averaged wind (Powell et al. 1991). Longer averaging of the flight level winds would tend to underestimate the true maximum 1-min wind speeds because the aircraft does not remain long in the peak gradient region on a radial flight track, especially in relatively small hurricanes like Andrew. The best method for adjusting these winds to the surface had previously been unclear, because the most thorough study of flight-level winds to surface wind observations (i.e., Powell and Black 1990) contained mainly tropical storm-force maximum winds and few observations in the eyewall region."
In discussions they usually seem to go with the 10 second data. (advisory data is supposed to be 1 minute sustained) I figured they just assumed they didn't sample the highest winds so they go with the value that is usually a bit higher since it is averaged over a shorter period. I just never saw it written anywhere that they do that until now.
----------------------------------------------------------------------------
Below is a snippet of an AVAPS file which goes into the ASPEN program to form a WMO standard dropsonde message. (UZNT13 for the Atlantic) It could have a thousand or more data lines, I only included 5 in the data part of it. (I don't know how to use the program much because there are other remarks, like the 61616 and 62626 sections which I don't know how you can get it to create in ASPEN) All this raw data is processed and adjusted, ignoring errors, and then you get the simpler dropsonde message text that comes across.
AVAPS-T01 STA 102125104 110827 094719.33
AVAPS-T01 COM UTC UTC Air Air Rel Wind Wind Vert GPS GPS Geopoten GPS Sonde Sonde GPS Wind GPS
AVAPS-T01 COM Sonde Date Time Press Temp Humid Dir Spd Veloc Longitude Latitude Altitude Wnd RH1 RH2 Snd Error Altitude
AVAPS-T01 COM ID yymmdd hhmmss.ss (mb) (degC) (%) (deg) (m/s) (m/s) (deg) (deg) (m) Sat (%) (%) Sat (m/s) (m)
AVAPS-T01 COM --------- ------ --------- ------- ------ ------ ------ ------ ------ ----------- ---------- -------- --- ------ ------ --- ----- --------
AVAPS-D01 S00 102125104 110827 095435.00 823.75 17.89 93.62 9.37 41.41 -9.87 -77.838640 34.095947 1497.49 9 93.62 92.93 9 0.62 1458.49
AVAPS-D01 S10 102125104 110827 095435.25 9999.00 99.00 999.00 9.10 41.50 -9.83 999.000000 99.000000 99999.00 9 999.00 999.00 9 0.61 99999.00
AVAPS-D01 S00 102125104 110827 095435.50 824.13 17.92 93.64 9.38 41.41 -9.73 -77.838676 34.095763 1493.47 9 93.64 92.97 9 0.62 1453.81
AVAPS-D01 S10 102125104 110827 095435.75 9999.00 99.00 999.00 8.61 41.41 -9.49 999.000000 99.000000 99999.00 9 999.00 999.00 9 0.62 99999.00
AVAPS-D01 S00 102125104 110827 095436.00 824.72 17.95 93.64 8.42 40.95 -9.53 -77.838711 34.095580 1487.34 9 93.64 92.97 9 0.63 1449.15
AVAPS-T01 COM Data Type/Data Channel: AVAPS SOUNDING DATA, Channel 1
AVAPS-T01 COM Project Name/Mission ID: Hurricane Irene, 20110827I1
AVAPS-T01 COM Aircraft Type/ID: NOAA P3, N43RF
AVAPS-T01 COM Launch Time (y,m,d,h,m,s): 2011/08/27, 09:53:33
AVAPS-T01 COM Sounding Name: drop 1
AVAPS-T01 COM Sonde ID/Type/Rev/Built/Sensors: 102125104, 1, A2, 2010/06/11 11:12, Vaisala RSS904, Ublox TIM-5H
AVAPS-T01 COM Sonde Freq/Batt/Firmware/Shutoff: 405.56 MHz, 8.0 v, 1.02, 32768 sec,
AVAPS-T01 COM Sonde Baseline Errors (p,t,h1,h2): -0.1 mb, 0.0 C, 0.0 %, 0.0 %
AVAPS-T01 COM Sonde Dynamic Errors (p,t,h): -0.4 mb, 0.0 C, 0.0 %
AVAPS-T01 COM Pre-launch Obs Data System/Time: IWGADTS Format (IWG1), 09:53:33
AVAPS-T01 COM Pre-launch Obs (p,t,d,h): 750.9 mb, 15.2 C, 15.5 C, 100.0 %
AVAPS-T01 COM Pre-launch Obs (wd,ws): 14.2 deg, 29.3 m/s
AVAPS-T01 COM Pre-launch Obs (lon,lat,alt): -77.833969 deg, 34.114975 deg, 2290.5 m, (077 50.0381'W, 34 06.8985'N)
AVAPS-T01 COM Operator Name/Comments: sans souci, lost data at 2400 ft
AVAPS-T01 COM Standard Comments: Weak Telemetry
AVAPS-T01 VER 3.1.1 SOFTWARE VERSION 3.1.1 - 2011 JAN 24; ADDS:
AVAPS-T01 FMT NOAA 1.7 FORMAT VERSION NOAA 1.7 - 2010 APR 17; ADDS: 0.25 SECOND WINDS
AVAPS-T01 TOF 0.00 Met/Wind Offset: ptu data leads wind data by 0.00 sec
AVAPS-T01 END 102125104 110827 095832.20
|