observational gaps - where would you fix if you could

[Jim Marusak]

something I know we most of the time have debated at one point, what if we had an observation here. or what if we had a sounding there. or wish the radar coverage wasn't so crappy there. Yes I know we also have complaints about computer power not coming up to speed fast enough for the models, wanting all our models running 4dvar instead of 3d, etc. but in this thread, this is more for real-time observations to get the right starting point to begin with.

 

so, if you had the chance, as the head of NOAA, to close those gaps  and make for a much better observational network, and were told by some hypothetical congress you would be given enough money to close them, but you just have to list them in a written down list for us to pay for them, what would you list?

 

my starter list:

 

 

upper air network :

 

one located at either New Braunfels, Austin (Bergstrom or Camp Mabry), or even Fort Hood/Killeen

the former CFB-Shilo site in Manitoba, or even Brandon

one near Saskatoon

 

radar:

central South Dakota near Pierre

west-central Minnesota near Alexandria

north central MN near Brainerd and/or International Falls

 

satellite:

get the new GOES-R satellites up and operational  ASAP

re-up funding for the polar satellite observing program

 

surface:

get all ASOS FAA sites equipped with sensors to see cloud cover up to 25-30kft, not just the current 12kft agl, and get them running asap

increase manual observers where feasible

 

let's see what everyone else could add to it and agree to.

[radarman]

something I know we most of the time have debated at one point, what if we had an observation here. or what if we had a sounding there. or wish the radar coverage wasn't so crappy there. Yes I know we also have complaints about computer power not coming up to speed fast enough for the models, wanting all our models running 4dvar instead of 3d, etc. but in this thread, this is more for real-time observations to get the right starting point to begin with.

 

so, if you had the chance, as the head of NOAA, to close those gaps  and make for a much better observational network, and were told by some hypothetical congress you would be given enough money to close them, but you just have to list them in a written down list for us to pay for them, what would you list?

 

my starter list:

 

 

upper air network :

 

one located at either New Braunfels, Austin (Bergstrom or Camp Mabry), or even Fort Hood/Killeen

the former CFB-Shilo site in Manitoba, or even Brandon

one near Saskatoon

 

radar:

central South Dakota near Pierre

west-central Minnesota near Alexandria

north central MN near Brainerd and/or International Falls

 

satellite:

get the new GOES-R satellites up and operational  ASAP

re-up funding for the polar satellite observing program

 

surface:

get all ASOS FAA sites equipped with sensors to see cloud cover up to 25-30kft, not just the current 12kft agl, and get them running asap

increase manual observers where feasible

 

let's see what everyone else could add to it and agree to.

 

SE Oklahoma and adjacent areas of NE TX is the worst radar gap in the country IMO.  Dangerous spot to lack coverage.  There is also a need for radar coverage in Wyoming coal region.    Brainerd as you mentioned is a good candidate as well.

[dtk]

something I know we most of the time have debated at one point, what if we had an observation here. or what if we had a sounding there. or wish the radar coverage wasn't so crappy there. Yes I know we also have complaints about computer power not coming up to speed fast enough for the models, wanting all our models running 4dvar instead of 3d, etc. but in this thread, this is more for real-time observations to get the right starting point to begin with.

 

so, if you had the chance, as the head of NOAA, to close those gaps  and make for a much better observational network, and were told by some hypothetical congress you would be given enough money to close them, but you just have to list them in a written down list for us to pay for them, what would you list?

 

my starter list:

 

 

upper air network :

 

one located at either New Braunfels, Austin (Bergstrom or Camp Mabry), or even Fort Hood/Killeen

the former CFB-Shilo site in Manitoba, or even Brandon

one near Saskatoon

 

radar:

central South Dakota near Pierre

west-central Minnesota near Alexandria

north central MN near Brainerd and/or International Falls

 

satellite:

get the new GOES-R satellites up and operational  ASAP

re-up funding for the polar satellite observing program

 

surface:

get all ASOS FAA sites equipped with sensors to see cloud cover up to 25-30kft, not just the current 12kft agl, and get them running asap

increase manual observers where feasible

 

let's see what everyone else could add to it and agree to.

 

Up funding and support for polar orbiters (MW and hyperspectral IR Sounders)....continued investment in better exploitation of current observing system (cloudy/precipitating fields of view), more emphasis on things like soil moisture, snow depth (SWE), vegetation, etc., as we push towards more fully coupled modeling systems

 

Fully fund Cosmic-2 (including second phase)

 

Space-based wind lidar (assuming that ADM-Aeolus is a success)

 

Launch a satellite with a wind scatterometer A S A P to supplement ASCAT

[Chinook]

The biggest radar holes in the Plains/Midwest are circled here. The most obvious is SE Colorado to OK Panhandle. Filling in the holes in other areas such as SE Oklahoma and NE Missouri would help more citizens in areas where dangerous weather affects more citizens.

 

[isohume]

surface:

get all ASOS FAA sites equipped with sensors to see cloud cover up to 25-30kft, not just the current 12kft agl, and get them running asap

 

The new sensors (CL31s) are already installed at most ASOS sites and all LCD sites. These new sensors detect clouds from the sfc to 25 Kft. Currently, the data is stored in ASOS and not sent out via METARs while ongoing side by side evals with the legacy (CL12K) ceilometers and upgrades to all the software packages are being performed.

[The_Global_Warmer]

How expensive is a Doppler radar?

[huronicane]

When you include civil works, a minimum of $1M

[huronicane]

I'd love to see every NWS office launch a sonde and fill in the blanks in more geographically expansive CWAs with an autosonde.

 

 

[radarman]

When you include civil works, a minimum of $1M

 

For an X band radar you are looking at about 400-500K depending on the siting.  C band maybe 600-700K.   There are ongoing costs as well of course.

[icebreaker5221]

The biggest radar holes in the Plains/Midwest are circled here. The most obvious is SE Colorado to OK Panhandle. Filling in the holes in other areas such as SE Oklahoma and NE Missouri would help more citizens in areas where dangerous weather affects more citizens.

 

Just how bad the coverage is out west really stands out, but it's actually much worse than this because this map doesn't account for beam blockage. Of course the need to make timely decisions based upon radar aren't nearly as common or urgent over the Rockies where tornadoes / sever thunderstorms are comparatively rare and when they do occur tend to be on the weaker side.  But it would be especially useful to fill the gaps to detect flash flooding or potentially help hikers avoid lightning in remote parks. 

 

Better radar coverage over the Rockies would also be nice for winter weather, but probably not a game changer as most of those decisions are made days in advance based primarily on model data, with perhaps occasional last-min upgrades to warnings based more upon surface obs than anything else. 

[isohume]

How expensive is a Doppler radar?

 

At the time, around 3M per site.

[mempho]

The biggest radar holes in the Plains/Midwest are circled here. The most obvious is SE Colorado to OK Panhandle. Filling in the holes in other areas such as SE Oklahoma and NE Missouri would help more citizens in areas where dangerous weather affects more citizens.

So, to sum it up, there is no coverage in either Area 51 or the Marfa region? Lulz.

[Jim Marusak]

from the radar map seen earlier in this post, I have to admit, there are a few surprising potential gaps. One being from north NJ to Allentown to Reading and Harrisburg; and the one in central NC.

 

as for the ones near the Canadian border and Sault Ste Marie, some of those gaps can be considered filled if you take into account the Canadian doppler coverage as well, especially near Sault Ste Marie. But what would help up there is if EC got off their high horse and made the full package available for distribution like we do in the states, especially when they finish the upgrades to dual-pol.

 

as for the radiosondes huronicane, are you suggesting maybe merge the two sets (current sites and the CWA's), or just a plain switch to the CWA's, as well as the others to fill the gaps (like INL)?

 

finally isohume, what's the timetable on the testing for the new sensors? any indication when they'll be ready to be published via metars?

[U_Thant]

Well, I dream of 88Ds lined up and down the eastern Bahamas (Great Abaco, Eleuthera, San Salvador, Mayaguana, and Grand Turk).

 

 

But that's just the             in me....

[The_Global_Warmer]

At the time, around 3M per site.

And that's startup ?

Not including maintenance

[wxdude64]

from the radar map seen earlier in this post, I have to admit, there are a few surprising potential gaps. One being from north NJ to Allentown to Reading and Harrisburg; and the one in central NC.

 

as for the ones near the Canadian border and Sault Ste Marie, some of those gaps can be considered filled if you take into account the Canadian doppler coverage as well, especially near Sault Ste Marie. But what would help up there is if EC got off their high horse and made the full package available for distribution like we do in the states, especially when they finish the upgrades to dual-pol.

 

as for the radiosondes huronicane, are you suggesting maybe merge the two sets (current sites and the CWA's), or just a plain switch to the CWA's, as well as the others to fill the gaps (like INL)?

 

finally isohume, what's the timetable on the testing for the new sensors? any indication when they'll be ready to be published via metars?

Yep, the areas north of Charlotte, the Allentown PA area, NW VA and eastern WV (plus LOTS of beam blockage from current radar to fill there too along with W NC and E TN) and something near Keene, NH to cover the east coast. Out west just needs help in general as far as radar coverage goes.

[Mallow]

Radar: Eugene/Bend, OR. Over 500k people in those two metro areas alone. Well within the darkest red on the above map.

[Chinook]

I'd love to see every NWS office launch a sonde and fill in the blanks in more geographically expansive CWAs with an autosonde.

 

Would it be possible for Canada to improve the radiosonde network? How about thinking outside the box: what if one to three US Navy vessels could surface and launch a weather balloon a couple of times per day, 300 miles off the West Coast.

[isohume]

finally isohume, what's the timetable on the testing for the new sensors? any indication when they'll be ready to be published via metars?

 

There is no concrete timetable that I have found. Basically, our el-techs say the new ceilometers are truncating

at 12 kft, simulating the legacy system. The FAA doesn't want clouds above this level right now, and they will be

the driving force as to when METARs include higher level clouds. Future software builds will allow high level clouds

to be ingested into the models, however.

 

And that's startup ?

Not including maintenance

That's correct. Maintenance is always an ongoing process and funded differently.

[Eskimo Joe]

something I know we most of the time have debated at one point, what if we had an observation here. or what if we had a sounding there. or wish the radar coverage wasn't so crappy there. Yes I know we also have complaints about computer power not coming up to speed fast enough for the models, wanting all our models running 4dvar instead of 3d, etc. but in this thread, this is more for real-time observations to get the right starting point to begin with.

 

so, if you had the chance, as the head of NOAA, to close those gaps  and make for a much better observational network, and were told by some hypothetical congress you would be given enough money to close them, but you just have to list them in a written down list for us to pay for them, what would you list?

 

my starter list:

 

 

upper air network :

 

one located at either New Braunfels, Austin (Bergstrom or Camp Mabry), or even Fort Hood/Killeen

the former CFB-Shilo site in Manitoba, or even Brandon

one near Saskatoon

 

radar:

central South Dakota near Pierre

west-central Minnesota near Alexandria

north central MN near Brainerd and/or International Falls

OK panhandle

SE OK

Elkins, WV - for DC/Baltimore Corridor

Add terminal doppler at Ft. Indiantown Gap for Lancaster-Allentown corridor

 

satellite:

get the new GOES-R satellites up and operational  ASAP

re-up funding for the polar satellite observing program

 

surface:

get all ASOS FAA sites equipped with sensors to see cloud cover up to 25-30kft, not just the current 12kft agl, and get them running asap

increase manual observers where feasible

Have an AWOS station at every high school in the country.  Four reasons:  (1) denser network of surface obs, (2) school admins / OEM can see what conditions are across their district to make better decisions on closings/delays, (3) Science teachers can use the instruments to inspire the next generation of atmospheric scientists, tech students on instrumentation, (4) Fire Department can utilize the conditions for plume modeling of hazardous material spills and in their emergency response efforts.

 

let's see what everyone else could add to it and agree to.

[Turtle]

Satellites:

get the new GOES-R satellites up and operational ASAP

 

GOES-R is scheduled for launch in March 2016.  It will have huge increases in data, as well as a new sensor that will collect total lightning.  Once the satellite is in orbit, it will undergo a 12 month check out period, but it WILL be ON during this timeframe.  Once this is finished, it will remain ON while in storage orbit until either GOES-East or West reaches the end of orbital life.  Yes, this means that GOES-R data WILL be available (a 3rd satellite lif you will) while in storage.  GOES-S will follow late 2017 according to current schedule. 

 

There will be a firehose of new data coming down, along with new algorithms that are being developed and tested through the GOES-R Proving Ground at various centers and cooperative universities and organizations across the country.  One of the important things that has been part of the GOES-R development and implementation has been to ensure that ALL users will be able to use the new data coming down from the satellite from day 1 once it is turned on.  Along those lines, that have been providing MANY online training presentations.  Anyone can go to the GOES-R website for the plethora of information available, and even see a simulated launch of GOES-R (under the Multimedia section).  The website is:

 

http://www.goes-r.gov

 

Be sure to go to the Training section of this website to see all the training available.  GOES-R training is also available through MetEd provided by COMET at UCAR in Boulder, CO.  Again, anyone can take this training once you sign up for a free account.  That site is:

 

https://www.meted.ucar.edu/

 

One more website with excellent pseudo GOES-R data already available.  As a part of the GOES-R Proving Ground, data has been collected from GOES-14 over the last few years, including several weeks each summer since 2012.  The data is being used while GOES-14 is in SRSOR (1-minute mode), which will be the new Rapid Scan mode on GOES-R, to save the data on whatever weather developed.  They managed to get a good variety of weather, from low clouds/fog to severe and fire weather, and are using this data for evaluation purposes.  Anyone, some of this data is available through CIMMS at the University of Wisconsin in Madison.  You can check this out at:

 

https://cimss.ssec.wisc.edu/goes_r/proving-ground.html

 

If you check out the CIMSS Satellite Blog (under the "Related Links" section), you can see near real time "interesting" information.  A few days ago, they posted updated IR satellite photos of Typhoon Dolphin from the new Himawari-8 (Japan/Pacific) satellite.  This blog covers MOST satellites, including the polar orbiters (VIIRS and Suomi-NPP).  Be sure to click on the GOES-R link and scroll down to see the previous GOES-14 data.

 

Having trouble finishing my post, so see Part 2 below... Turtle

[Turtle]

Here's part 2, since it seems that I can't add more to my previous post...

 

So, how do I know all this stuff??  I am the satellite focal point for my office.  Not all NWS offices have a satellite focal point, but I am here.  I have had some incredible opportunities over the last couple of months to learn about and help in the development of GOES-R products and the new algorithms, and will be doing one more thing in a couple of weeks.  

 

Late last year, I was selected by my regional office to take part on a national evaluation team for the GOES-R Proving Ground at the NWS Training Center in Kansas City.  I went out for a week in March with two other NWS field forecasters to evaluate simulated GOES-R data in 8 simulations, using data from past events in 2013 and 2014, to help out in determining the usefulness of the new algorithms and capabilities (added data) that GOES-R will provide.  It was a pretty amazing week!

 

Then, I attended the 2015 NOAA Satellite Conference in Greenbelt, MD, at the end of April.  Over 600 people attended from around the world!  I learned SO much information about the worldwide satellite cadre, including Himawari and the polar orbiters, and how the U.S. and many other countries are working together to make sure that satellite data will continue to flow for years to come.  You can see the presentations that were given at the NSC at their website:

 

http://satelliteconferences.noaa.gov/2015

 

Click on the Presentations Link, then read through the schedule.  I realize that some of the presentations have "boring" titles, especially in Section 1.  Some of those were very interesting, though.  If you want to see a show, just click on the Session number.  It should be a hyperlink with the presentation.

 

One more trip for me.  In a couple of weeks, I will be taking part in the Hazardous Weather Testbed, 2015 Spring Weather Experiment at SPC in Norman.  GOES-14 will be turned on, again to gather data in SRSOR (1-minute) format.  We will be evaluating data in real time, and hoping Mother Nature cooperates and gives us some good (and varying) opportunities to test out the algorithms and added data.  Will see what happens during the first week of June!!

 

If you've got any questions, PM me!

 

--Turtle  

 

[jgf]

what's the use of more observations, if they don't use the ones they have?

 

looking mostly at 10m wind speed and direction, i often see that  the f00 for the models i look at is radically different than the observed winds at that time  and location.

 

i know that the f00 is not the initialization - we had a thread a bout it - but you would think it would be pretty close. certainly the f00 is very similar to the anal file...

[Mallow]

what's the use of more observations, if they don't use the ones they have?

 

looking mostly at 10m wind speed and direction, i often see that  the f00 for the models i look at is radically different than the observed winds at that time  and location.

 

i know that the f00 is not the initialization - we had a thread a bout it - but you would think it would be pretty close. certainly the f00 is very similar to the anal file...

 

Wind speed and direction can be extremely sensitive to local effects, so even if there's an observation at a station that shows a particular wind speed/direction, that doesn't mean it's representative of the larger-scale flow. Moreover, I don't know what product you're looking at in particular, but many model visualizations show wind vectors at widely-spaced intervals, and on those products, it's not particularly common for a wind vector to happen to be right near an ASOS. If you're referring to non-ASOS stations, wind data from less well-controlled stations are notoriously unreliable. Random errors and biases for wind data are generally much worse than for variables like temperature and even humidity.

 

The idea that "observations aren't used" in model analysis and initialization is mistaken--the data are ingested and utilized to the best of our current abilities.