USGS Rapid Release Gauge Dataset

[kwilcox]

@rsignell-usgs and I were discussing some issues with this data-set and decided to summarize what we found here so others could comment and offer insight.

Datafiles are available here: http://ga.water.usgs.gov/flood/hurricane/sandy/datafiles/

Water Level files

Files that end in *WL.txt and have # Site type = water level in the metadata header.
Example

1. There may be multiple cases (I only found one in my limited analysis) where two stations are measuring the same parameter at the same location, but returning different elevation values. This brings up the question of what the values are actually representing. Is it water level above NAVD88, or water level above the sensor? Subtracting the sensor height in either case gives equal values for elevation for both stations.

   • http://ga.water.usgs.gov/flood/hurricane/sandy/datafiles/SSS-CT-MSX-019WL.txt

     ...
     # Sensor location latitude 41.281111        
     # Sensor location longitude -72.352222
     # Sensor elevation above NAVD 88 = 1.72 ft
     ...
     date_time_GMT   elevation   nearest_barometric_sensor_psi
     10-29-2012 10:00:00 1.82    14.3967
     10-29-2012 10:00:30 1.83    14.3944
     10-29-2012 10:01:00 1.83    14.3967
     

   • http://ga.water.usgs.gov/flood/hurricane/sandy/datafiles/SSS-CT-MSX-020WL.txt

     ...
     # Sensor location latitude 41.281111        
     # Sensor location longitude -72.352222
     # Sensor elevation above NAVD 88 = 2.17 ft
     ...
     date_time_GMT   elevation   nearest_barometric_sensor_psi
     10-29-2012 10:00:00 2.27    14.3967
     10-29-2012 10:00:10 2.27    
     10-29-2012 10:00:20 2.26    
     10-29-2012 10:00:30 2.28    14.3944
     10-29-2012 10:00:40 2.28    
     10-29-2012 10:00:50 2.27    
     10-29-2012 10:01:00 2.27    14.3967
     

Wave files

Files that end in *WV.txt and have # Site type = wave height in the metadata header.
Example

1. There is no indication of what the values actually represent in the metadata header other than Site type = wave height
2. It appears to be a high frequency elevation reading (every 2 seconds), so the same issues may apply Water Level section.

Pressure files

Files that end in *BP.txt and have # Site type = barometric pressure in the metadata header.
Example

1. There is no indication in the metadata header what unit the values are in. After some email traffic, @rsignell-usgs discovered they are in psi.
2. The column name is elevation

[rsignell-usgs]

@hjenter, can you review this and help us address the lingering issues? Or point us to who we should talk to?

[hjenter]

Without contacting the original processor of the data, I cannot be sure. However, my guess is that there is an error in the datum conversion for one or both of the co-located sensors in your Water Level files example. I'm fairly certain that the intent was to provide water surface elevation above NAVD88.

Water Science Centers occasionally deploy two pressure loggers at the same location for redundancy. This is probably the reason for two files of the same parameter at the same location. It appears that, in your example, one logger was logging every 30 seconds and one was logging every 10 seconds. Perhaps, they were just trying to determine what information the 30-second logger was missing.

Your write-up for the Wave files is correct. They are simply more-frequently-sampled water levels.

Your write-up for the Pressure files is correct. They are incorrectly labeled "elevation" when they should be labeled "air pressure, psi"

[rsignell-usgs]

@hjenter, thanks for checking this.

Who should we report the likely error in the data conversion to?
I pasted the lat/lon from the header (41.281111, -72.352222) into google search (I love that I can do that) and it pops up this map for Saybrook CT, so someone at USGS in CT?

[hjenter]

Contact for CT:
John R. Mullaney
Hydrologist
U.S. Geological Survey
New England Water Science Center
Connecticut Office
101 Pitkin Street
East Hartford, CT 06108

[rsignell-usgs]

Awesome. I'll contact John via e-mail and hopefully we can get this figured out.

[jmusgs]

Ok, I sent an email to help resolve this. The short answer is the sensor for CTMSX20 was out of the water at your time of comparison of the two files shown above. This sensor was installed at a slightly higher elevation than CTMSX19. I would be careful in comparisons, as many of the sensors for this storm have periods of data where sensors were out of the water. these are represented by flat lines approximately at the lowest record-able water level. Hope this helps!

[jmusgs]

graphic to explain above comment

[hjenter]

Unvented sensors which are not submerged at all times are very common for
Water Science Center storm surge deployments. Water Science Centers often
deploy sensors in locations that are expected to be inundated during a
storm, but are not inundated normally. In addition, Water Science Centers
almost exclusively deploy unvented pressure sensors without the use of
boats. Therefore, deployment locations tend to be right at the water's
edge, implying that many of the sensors come out of the water at low tide.
The main purpose of these sensors is often to track the peak storm tide and
there is less concern for capturing the entire time period of deployment.
As the new pressure sensors become more prevalent, efforts likely will be
made to deploy the sensors in locations such that the instrument is
submersed at all times during future storms.

=-=-=-=-=-=-=-=-=-=-=-=
Harry L. Jenter, PhD
Deputy Chief, Office of Surface Water
USGS
12201 Sunrise Valley Drive
Mailstop 415, National Center
Reston, VA 20192
703 648-5916, office
703 295-2800, cell

On Fri, Mar 4, 2016 at 11:26 AM, jmusgs [email protected] wrote:

Ok, I sent an email to help resolve this. The short answer is the sensor
for CTMSX20 was out of the water at your time of comparison of the two
files shown above. This sensor was installed at a slightly higher elevation
and CTMSX19. I would be careful in comparisons, as many of the sensors for
this storm have periods of data where sensors were out of the water. these
are represented by flat lines approximately at the lowest record-able water
level. Hope this helps!

—
Reply to this email directly or view it on GitHub
#2 (comment).