| 1 | From: Stig Syndergaard [ssy@dmi.dk]
|
|---|
| 2 | Sent: 10 February 2015 12:06
|
|---|
| 3 | To: Burrows, Chris
|
|---|
| 4 | Cc: Culverwell, Ian
|
|---|
| 5 | Subject: Re: ROPP_PP
|
|---|
| 6 |
|
|---|
| 7 | Hi Chris,
|
|---|
| 8 |
|
|---|
| 9 | I agree with your observations for the lat, lon, azimuth interpolation in the code, and from
|
|---|
| 10 | your figures it looks like you've found a good solution. Why only temporary?
|
|---|
| 11 |
|
|---|
| 12 | No doubt that the occ_tool (and the invert_tool) has some issues. Maybe we should
|
|---|
| 13 | make a coordinated effort to get them solved for the next ROPP release. We have the
|
|---|
| 14 | ones from Kjartan, my Struct.Uncert. report, our DMI tickets
|
|---|
| 15 | (https://trac.romsaf.org/ropp/report/10) and now your investigations with GNOS -
|
|---|
| 16 | probably with some overlap.
|
|---|
| 17 |
|
|---|
| 18 | I also think we should aim at merging the occ_tool and the invert_tool (I think I've
|
|---|
| 19 | mentioned that earlier), it would make maintenance easier.
|
|---|
| 20 |
|
|---|
| 21 | Ian: Would it be possible to schedule all this for ROPP 9?
|
|---|
| 22 |
|
|---|
| 23 | -Stig
|
|---|
| 24 |
|
|---|
| 25 | On 2015-02-05 14:38, Burrows, Chris wrote:
|
|---|
| 26 | > Many thanks Stig,
|
|---|
| 27 | >
|
|---|
| 28 | > That report is very helpful indeed - it's good to know that you have seen the same
|
|---|
| 29 | features (e.g. Fig 2.5). I've come up with a temporary fix for the lat/lons (see attached),
|
|---|
| 30 | though it's not a robust solution.
|
|---|
| 31 | >
|
|---|
| 32 | > Please don't let this distract you from your other work, but any input would be gladly
|
|---|
| 33 | received!
|
|---|
| 34 | >
|
|---|
| 35 | > Thanks again
|
|---|
| 36 | >
|
|---|
| 37 | > Chris
|
|---|
| 38 | >
|
|---|
| 39 | >
|
|---|
| 40 | >
|
|---|
| 41 | > -----Original Message-----
|
|---|
| 42 | > From: Stig Syndergaard [mailto:ssy@dmi.dk]
|
|---|
| 43 | > Sent: 05 February 2015 13:18
|
|---|
| 44 | > To: Burrows, Chris
|
|---|
| 45 | > Cc: Culverwell, Ian
|
|---|
| 46 | > Subject: Re: ROPP_PP
|
|---|
| 47 | >
|
|---|
| 48 | > Hi Chris, Ian,
|
|---|
| 49 | >
|
|---|
| 50 | > I'll have a look at this over the next couple of days (into next week); today I am just
|
|---|
| 51 | swamped in other work. I remember from the SU project some years ago that the lat, lon
|
|---|
| 52 | vector was a bit odd. And also a kink at 25 km in std.dev. I've attached the report that I
|
|---|
| 53 | did back then. At the end (section 3) there is a summary of the issues that I found. It's
|
|---|
| 54 | been on my mind to fix these things in ROPP for a while, but I never seem to get around
|
|---|
| 55 | to do it (always something more urgent).
|
|---|
| 56 | >
|
|---|
| 57 | > Best,
|
|---|
| 58 | > -Stig
|
|---|
| 59 | >
|
|---|
| 60 | > On 2015-02-05 13:19, Burrows, Chris wrote:
|
|---|
| 61 | >> Hi Stig, Ian,
|
|---|
| 62 | >>
|
|---|
| 63 | >> Just a quick update;
|
|---|
| 64 | >>
|
|---|
| 65 | >> It seems that the IPs are calculated for GO, and then, for the WO processing, the IPs
|
|---|
| 66 | below 25km are overwritten with uniformly spaced levels (keeping the size of the array
|
|---|
| 67 | the same). The lats, lons and azimuths *aren't* mapped onto these levels, so the final
|
|---|
| 68 | interpolation
|
|---|
| 69 | (https://trac.grassaf.org/ropp/browser/ropp_src/trunk/ropp_pp/tools/ropp_pp_occ_tool.f90
|
|---|
| 70 | #L731 ) is mismatching lat,lon,azim with the post-WO impact parameters. Maybe we
|
|---|
| 71 | should store the pre-WO values for this interpolation step. It's not a huge deal, but would
|
|---|
| 72 | be important for TPD in assimilation, and 2D operators.
|
|---|
| 73 | >>
|
|---|
| 74 | >> As for the jump in standard deviations, I wonder if it's related to the fact that the WO
|
|---|
| 75 | processing outputs IPs on a fine grid (~5m in the case I was looking at). This is then
|
|---|
| 76 | interpolated onto a coarse grid, so I fear that a lot of information is being lost (this is
|
|---|
| 77 | probably consistent with large noise but small correlations).
|
|---|
| 78 | >>
|
|---|
| 79 | >> I'll keep looking into this, but I'm sure I'm not qualified to mess around with the WO
|
|---|
| 80 | settings - scary stuff...
|
|---|
| 81 | >>
|
|---|
| 82 | >> Thanks
|
|---|
| 83 | >>
|
|---|
| 84 | >> Chris
|
|---|
| 85 | >>
|
|---|
| 86 | >>
|
|---|
| 87 | >>
|
|---|
| 88 | >> -----Original Message-----
|
|---|
| 89 | >> From: Burrows, Chris
|
|---|
| 90 | >> Sent: 04 February 2015 18:28
|
|---|
| 91 | >> To: Stig Syndergaard
|
|---|
| 92 | >> Cc: Culverwell, Ian
|
|---|
| 93 | >> Subject: ROPP_PP
|
|---|
| 94 | >>
|
|---|
| 95 | >> Hi Stig,
|
|---|
| 96 | >>
|
|---|
| 97 | >> I was looking at some sample data from the Chinese GNOS instrument with Ian today
|
|---|
| 98 | - they are using ROPP for the preprocessing. It looks pretty good but there is a sharp
|
|---|
| 99 | jump in the BA O-B standard deviations at 25km (see attached O-B stats and vertical
|
|---|
| 100 | correlations). This presumably is related to the WO/GO transition (Kjartan made
|
|---|
| 101 | reference to something similar here: https://trac.grassaf.org/ropp/ticket/293).
|
|---|
| 102 | >>
|
|---|
| 103 | >> We did a bit of further digging and what we found may or may not be related, but it is
|
|---|
| 104 | certainly confusing....
|
|---|
| 105 | >>
|
|---|
| 106 | >> In the output file, the impact parameters are spaced equally, but there are sudden
|
|---|
| 107 | jumps in the tangent point lat-lons corresponding to IH=25km. The values do correspond
|
|---|
| 108 | to a sensible track; the transition is just due to a change in the data sampling. The
|
|---|
| 109 | azimuth also shows a suspicious jump. These are plotted in the attached 'coords.png'
|
|---|
| 110 | file (note that the jagged steps in the azimuth are due to BUFR precision; they are not
|
|---|
| 111 | present in the full-resolution files, though the 25km feature is still there).
|
|---|
| 112 | >>
|
|---|
| 113 | >> The bottom right plot is the 'angular distance', sqrt((delta lat)*2)+(delta lon)^2)) as a
|
|---|
| 114 | function of impact height (i.e. a vertical cross section of the TP trajectory) - this looks very
|
|---|
| 115 | suspicious!
|
|---|
| 116 | >>
|
|---|
| 117 | >> Similar plots are produced with GRAS data using ROPP, so we suspect there may be
|
|---|
| 118 | a bug. Do you have any thoughts about this? Perhaps the lat/lons are being overwritten
|
|---|
| 119 | incorrectly for the WO part of the processing (still sampled against time rather than a
|
|---|
| 120 | fixed IP grid?).
|
|---|
| 121 | >>
|
|---|
| 122 | >> Thanks in advance for any comments, (please copy Ian into a reply
|
|---|
| 123 | >> too!)
|
|---|
| 124 | >>
|
|---|
| 125 | >> Chris
|
|---|
| 126 | >>
|
|---|
| 127 | >>
|
|---|
| 128 | >> ---
|
|---|
| 129 | >> Chris Burrows Satellite Applications Scientist
|
|---|
| 130 | >> Met Office FitzRoy Road Exeter EX1 3PB United Kingdom
|
|---|
| 131 | >> Tel: +44 (0)1392 88 6449
|
|---|
| 132 | >> E-mail: chris.burrows@metoffice.gov.uk http://www.metoffice.gov.uk
|
|---|
| 133 | >>
|
|---|
| 134 | >>
|
|---|
| 135 | >
|
|---|
| 136 |
|
|---|
| 137 | --
|
|---|
| 138 | _________________________________________
|
|---|
| 139 | Stig Syndergaard, PhD
|
|---|
| 140 | Danish Meteorological Institute
|
|---|
| 141 | Lyngbyvej 100
|
|---|
| 142 | DK-2100 Copenhagen
|
|---|
| 143 |
|
|---|
| 144 | Phone: +45 39 157 408
|
|---|
| 145 | E-mail: ssy@dmi.dk
|
|---|
| 146 | _________________________________________
|
|---|
