GRUAN operations: Suggestion to do regular parallel ascents

Franz Immler wrote:

During the Lindenberg meeting it was agreed that among other things GRUAN station were required to conduct

1 x weekly production radiosonde with the best technology currently available (GCOS-121)

this suggestion has already often been argued about and is certainly one of the weaker points in the current concept of GRUAN. It obviously refers to an commercially available standard radiosonde which is currently launched at most sites 2 or even 4 times a day anyway. Even though some plans referring to extended ground check and alike are being discussed, it is currently not clear what should be the benefit of the additional launch of such an instrument.

What I suggest therefore, is to launch (at least) once a week two different high quality radiosondes on one balloon. Every GRUAN station can select one additional type of radiosonde to what they regularly start anyway and conduct a parallel launch of the two radiosondes once a week, better twice a week, one day and one night time. In principal every station can choose any radisonde they like but it is recommended to seek guidance from the lead centre.

The benefits with respect to the GRUAN goals of such regular parallel launches are the following:

  • The parallel launch immediately provides the required redundancy of measurements.
  • Based on the comparison of the two radiosonde profiles, stringent QA/QC is a straight forward task. Issues with one of the instruments can more easily be identified.
  • Each of the sensors of the sondes may be advantageous in different regimes (e.g. altitudes), allowing for the retrieval of a higher quality product by merging the profiles of the two sensors. If both sensors are biased due to the same source (e.g. radiation) but with different strength, the differences observed between the two signals indicates the strength of the source and can be used as an input parameter for correction schemes.
  • In case that unannounced technical changes are introduced by the manufacturer of one of the radiosondes, its influence on the data quality can be assessed based on the comparison to the other sonde in retrospect. It is rather unlikely that such changes occur for both types at the same time.
  • “Managed change”: If a station regularly operating radiosondes of type A and B plans to replace type A by another type, say C, managed change is a straight forward task since basically all it has to do is to make sure that type B sonde is stable during an sufficiently long period of time (e.g. one year) before and after the change from A to C. This can be ensured by putting enough radiosondes of one batch in stock. In case A is the operational radiosonde launched twice daily, the first step of the change would be to use B for the daily launches and A for the weekly reference launch and after a while replace A by C and later use C for the daily soundings. Some additional triple launches should be performed if feasible.

Concerning the homogeneity of the GRUAN network the double launch plan is very helpful for the following reasons:

  • A large subset of GRUAN stations will share at least one of type of radiosonde and therefore directly provide an homogeneous dataset. In case two station do not share a common type (station I launches types A and B while Station II launches C and D) a homogeneous dataset can be achieved using the statistics of a GRUAN station that launches, for instance, the types A and C.
  • Based on long term observations it will be possible to obtain a homogeneous dataset from all GRUAN station based on such comparison statistics, including the comparisons to the monthly reference sonde. Even though it will take quite some effort to do this, the advantage is that such a dataset will be more robust than a dataset from a network that used one single type of radiosonde only, since the latter will be a lot more vulnerable to technical issues with that particular radiosonde, its biases and to whatever changes occur with that radiosonde or its manufacturer.
  • GRUAN will build up a long term dataset of parallel launches of production radiosondes and also comparisons to the monthly reference type radiosondes. This dataset can do a great job for homogenizing the entire aerological dataset of GOS.

For the operation of a GRUAN station, the introduction of a second radiosonde is done fairly easy and for moderate costs. Depending on what equipment is already installed at the station, it can be done as easily as hooking another radio on the antenna of the existing sonde and acquiring the signal from the second sonde with one additional computer. The investment in new equipment can therefore be well below 10 k€. The cost for the additional 52 or 104 radiosondes a year should also manageable ( some 7 to 15k€ ). The cost for additional gas and larger balloon will not be that significant either. Again depending on what type of instruments are used, the additional effort to do a parallel instead of a single launch is not more than 1-2 personhours per launch, possibly even less.

2 Responses to GRUAN operations: Suggestion to do regular parallel ascents

  1. Peter Thorne says:

    I’m mulling over all the comments (a distinct advantage over everyone else) received to this section of the draft Imp plan. And its abundantly clear that there is no consensus at least in initial positions. Everyone wants to tweak these things but everyone wants to do it differently.

    At the heart of this issue is that there is no rigorous quantitative basis behind the 121 position or, for that matter, anybody else’s. Until we have that I don’t see how we can avoid endless arguments and attempts to tweak the list. All well intentioned but in the longer term they will potentially imperil GRUAN’s credibility.

    What I would therefore urge is quantitative study of the issue. This should include many things that Franz suggests and many others.

    This forward looking perspective with a view to revising the current guidance just the once and with a quantitatively defensible basis say in late 2011 is the most sensible route forwards in my view.

    How we get this quantitative basis will depend upon active engagement and experimentation at the sites (some along the lines Franz suggests, and others too including e.g. frequency of CFH soundings etc); engagement with GSICS / NASA on the overpass issue; and quantitative study using historical data / climate model output.

    The quantitative analysis would necessarilly include temperature and water vapour.

    Bottom line is that 121 is a minimum set and clearly not a long-term solution. Nor for that matter was it ever proposed as one. We need active experimentation and analysis to alight on a quantifiably defensible set of recommendations to supercede 121 with. This is high priority. Let us work towards that rather than go backwards and revisit what has been made clear is an interim measure?

  2. June Wang says:

    My first time to put some things here. It is “cool”. Today I started to look at what we want to do in order to address the topics of “Management of Change”. Like a lot of Franz’s suggestions, and totally agree with Peter that we need “quantitative basis” for the ideas. We will start to work on Lindenberg data to see whether we can provide some basis. Your suggestions would be more than welcome.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: