Thursday, September 17, 2015

New Type of SLC Detection Model - 4

Fig. 1 GeoZones "aa"-"br'
I have some good news about the new sea level rise (SLR) and sea level fall (SLF) software model.

The aggregation logic is working to combine all of the tide gauge stations in a geoZone ("aa" - "br") into two groups.

The two groups are either the SLF or SLR type, all aggregated into one of two mean seal level sub-groups.

The logic prior to this had only labelled a tide gauge station as either SLF or SLR.

Fig. 2
The model back then did no work on them, but now aggregation logic is in place, and I can graph any GeoZone on one graph, as an SLF track and an SLR track (Fig. 2 and Fig. 3).

It shows both the SLF and SLR over the years and shows SLF in red, SLR in black (Fig. 2, and Fig. 3).

Even better, I have the future projection software working with the aggregation logic, although it is not yet perfected (maybe tomorrow).

Fig. 3
Which means that the aggregation logic (after having aggregated all the historical sea level values for all tide gauge stations on the planet that are PSMSL into one array) can hand it off the the future projection logic.

Fig. 4 An "SLR only" GeoZone
The future projection logic uses that historical data, as well as the multi-value doubling logic, to project the SLR and/or SLF data out to the year 2100.

See Fig. 4, which among other things, shows where the future projection graph attaches.

It will all (historical & future) still be on one graph.

As you can only give a couple of examples today, but hang on, because some very interesting revelations are on the way.

A final note about PSMSL.

The values are in RLR millimeters, as you can see on the graphs, so read up on RLR millimeters if you haven't yet (About RLR millimeters).

Figure shows the basic idea, which is, RLR millimeters are 7000mm below global mean sea level (it is explained in the link I just gave).

As I understand it.

So in Fig. 4 there is a global mean sea level line at the 7000mm mark.

Anything above the 7000mm line is above the global mean sea level average, and anything below it is below the global mean sea level average.

Anyway, tomorrow I hope to print some graphs with the future SLF and SLR projections on it, as a continuance added to the end of the historical record.

In conclusion, I see the SLC GeoZones as a high level monitoring tool, the sub-zones SLF and SLR as closer monitoring, and when necessary, the individual tide gauge stations as the closest monitoring.

UPDATE: The SLR only scenario in GeoZone "aa" as  shown in Fig. 4 is anomalous, but I have not discerend all the reasons.

But one reason is that the zone cuts off part of East Greenland, and the zone east of zone "aa" is zone "af" which does have 5 SLF stations, 1 SLR station:

Zone: af (Lat: 60.000000 -> 90.000000) (Lon: -60.000000 -> -120.000000)
QIKIQTARJUAQ (Lat: 67.866669, Lon: -64.116669) (2009->2010) (6967->6961) (6964,SLF)
RESOLUTE (Lat: 74.683334, Lon: -94.883331) (1958->1976) (6982->6928) (6942,SLF)
LITTLE CORNWALLIS ISLAND (Lat: 75.383331, Lon: -96.949997) (1992->1993) (6993->7017) (7005,SLR)
ALERT (Lat: 82.489998, Lon: -62.320000) (1968->2007) (6988->6976) (6957,SLF)
CAMBRIDGE BAY (Lat: 69.116669, Lon: -105.066666) (1965->1973) (6993->6967) (6969,SLF)
ULUKHAKTOK ( FORMALLY HOLMAN ) (Lat: 71.233330, Lon: -118.266670) (2003->2009) (7047->6959) (6958,SLF)

So, I may have to change the zone lines to comport with the fingerprint boundaries more accurately, or something to that effect.

UPDATE 2: On closer analysis, the zone "af" stations don't resolve the problem because most of them are inactive now, and the records available for when they were active, are only for a few years:
(Alert)
year,rlrmm
1968,6988
1969,6880
1971,7077
1973,6981
2003,6940
2004,6913
2005,6929
2006,6943
2007,6976
(Qikiqtarjuaq)
year,rlrmm
2009,6967
2010,6961
(Little Cornwallis Is.)
year,rlrmm
1992,6993
1993,7017
(Resolute)
year,rlrmm
1958,6982
1959,7009
1962,6936
1963,7033
1964,7035
1966,7009
1968,6974
1969,6993
1971,7026
1972,6985
1974,6944
1975,6944
1976,6928
(Cambridge Bay)
year,rlrmm
1965,6993
1966,6982
1967,7028
1968,6960
1969,6963
1970,6946
1972,6986
1973,6967
That is not enough data ("minimum of 30 years" - Mitrovica), so I may have to get some satellite data to resolve those two zones ("aa" & "af").

3 comments:

  1. I don't mean to take up your time, Dredd, but could you explain how satellite data, which came along after the buoys were in place, can provide more information for your model? [by this very question you can see that I don't get it]

    (feel free to shoot me an e-mail and remove this comment)

    Tom

    ReplyDelete
    Replies
    1. The satellite records may not salvage all locations, if any.

      It depends on when those satellite records begin, and when the tide gauge records end.

      For example, Alert ends in 2007, so satellite records of sea level could be spliced after conforming them to rlrdata format.

      However, since Cambridge Bay, for example, ends in 1973, maybe it can't be reconciled.

      Some of the station data is like dinosaurs, but not so many that would prevent project completion.

      I have the future projector working now.

      I am going to clean up the zones and data this weekend.

      Mitrovica did not respond to emails about the golden tide gauge station sites (the 23 Douglas recommended, so I will keep as many as I can.

      I am sure it will be over 1,200 or 1,300 sites after taking out those with less than 30 years of records, and those that stopped recording more than ten years ago.

      Mitrovica sometimes has used satellite data that way sometimes.

      Anyway, good questions IMO.

      Delete
  2. I was wrong.

    There are only 728 tide gauge stations with more than 30 years or more of records.

    Can't change reality.

    But remember that Mitrovica et. al. only used the golden 23.

    ReplyDelete