Saturday, October 17, 2015

New Type of SLC Detection Model - 8

Fig. 1
I. Intro

I made an improvement in the model that I should have caught earlier.

Evidently, I got lazy doing the transition from actual historical data into future projection data.

Wanting to preserve the individual characteristics of each location around the globe, I
Fig. 2
did not do enough to make a meaningful transition from the past into the future.

Thus, if you look at the transition zone in the graphs, where the last historical year transitions into the first future year, it is askew, i.e., either radically high or radically low, in too many instances.

Fig. 3
So, I have altered the code in the C++ class, where the transition logic was located, concerning only sea level rise (SLR) locations.

The sea level fall (SLF) locations are not implicated, since they are in another C++ class and did not exhibit the problem.

II. Software Change Results

Today's graphs illustrate the modification results, which make for a better transition in sea level change (SLC) models.

In all the SLC graphs posted on Dredd Blog recently, I put a red "dot" where the
Fig. 4
change from actual historical sea level data transitions into future projected sea level.

Compare yesterday's graphs of the Philadelphia area (The Extinction of Philadelphia) with today's regenerated graphs that contain the improvements, and you will see what I mean.

Fig. 5
For example, there was an immediate jump in sea level shown at the transition position.

That event actually took place historically when the Philadelphia station was first brought on line, and since I was trying to keep the characteristics of each location going past the historical data and on into the future, it was mistakenly repeated at the front of the projection data.

One problem with that little hill or bump there is that it would cause deniers and detractors to immediately discount the entire projection because "see, that little jump in sea level didn't happen, so how can you trust the rest of it?"

III. Aesthetic Results

Now, the transition between the historical and future data follows the trend line
Fig. 6
instead, and the historical characteristics have been inversed.

Which means that the end of the historical records meets up with the beginning of the future data more smoothly by using the actual historical trend line existing at the time of the transition.

Further, the SLR / SLF characteristics of each station are still preserved and placed in the future projection data.

That is, where a station has a volatile up / down, SLF / SLR historical pattern, that characteristic will still show up in the future projection, preserving some of the nature of that station location.

IV. A Scientific Inquiry Emerges

This SLR / SLF saw-tooth pattern in graphs brings up an interesting question.

In order to understand it completely, we will first need to review some Dr. Mitrovica discussions.

Remember that, as the ice sheets melt or flow into the sea, the gravity of their mass that holds/held sea water close to the shore (a gravity caused very long lasting surge, like what onshore storm winds do when they create a temporary storm surge) will diminish, and so the sea level close to shore falls (SLF).

The freed-up water goes somewhere else to be part of SLR, but not only that, Greenland and Antarctica are both doing this (but not always in sync).

Greenland may cause some SLF at one location, while Antarctica may cause SLR at that same location (but not always in sync).

This would have an ongoing see-saw impact on tide gauge stations, depending on their proximity, or lack thereof, to each of those two ice sheets.

Heavily glaciated areas are also players in the see-saw generating pattern (see e.g. Proof of Concept - 3).

V. Conclusion

SLC is composed of both SLF and SLR at a global level, and at the local station level as well.

The graphs which completely smooth that over by way of using "global mean sea level average" tend to cover that reality up (we can thereby lose sight of the volatility of SLC).

They all do good work though, because the main thing to remember is that SLC, whether SLF or SLR, is a current threat to national security (Has The Navy Fallen For The Greatest Hoax?).

The next post in this series is here, the previous post in this series is here.

1 comment:

  1. Great catch Dredd! Your refinement, in line with assumed trend line, makes your model more robust, "legitimate" and probable. We'll need to see how it holds up over the remaining years.