The World According To Measurements - 8

Fig. 1 Are your measurements anomalous?

I. Introduction

One of the curses these days could be "may all of your measurements be anomalous" (The Damaged Global Climate System - 7).

The hurricanes Harvey, Jose and Irma are of the anomalous ilk, i.e., not part of an undamaged global climate system (ibid).

Where there are no measurements there are no anomalies, but neither are there any facts upon which to form an inkling of reality.

The measurements we collect and record are our only way of fashioning and forging the pillars, posts, and beams of our science reality.

II. From Measurements To Graphs

Today, I wanted to illustrate how our world depends on measurements, especially scientific climate / weather measurements, so as to form our coherent hypotheses, theories

Fig. 2a

and concepts of natural law (such as the law of gravity).

The graphs in today's post  are constructed from measurements taken the world over during a span of years exceeding a century, then stored in the WOD, PSMSL, GISS, and other datasets.

Those graphs compose a pair of contrasting views; a contrast that is formed by the use of 99.99% of PSMSL tide gauge records to form one graph, together with the use of a smaller set of only twenty three zones of PSMSL tide gauge records to form the other graph in the pair.

Fig. 2b

The smaller set of the pair is called the "Golden Twenty Three" (g23), which is a subset of the total tide gauge station records used in the full set.

The g23 are 23 PSMSL tide gauge stations that were thoughtfully and fairly selected by learned scientists.

They sought a fair representation of PSMSL tide gauge stations with which to form a robust view of global sea level characteristics over a long span of time.

Those twenty three representatives were selected because they work well to give a balanced (Fig. 1) set of measurements for pondering global sea level change (Golden 23 Zones Revisited).

I have expanded just a bit upon the g23 to make them the 23 WOD Zones.

I use all tide gauge stations in any zone which contains at least one of the golden twenty three individual tide gauge stations.

Thus, the g23 is composed of 25 individual tide gauge stations located in only 23 individual WOD zones (two of those zones have an additional tide gauge station).

The larger set that contrasts with the g23 is composed of 99.99% of all PSMSL tide gauge stations.

There are 1,486 total tide gauge stations in the database.

Four stations are excluded (#1975 "SANTANA", #1963 "IQUIQUE", #571 "TALCAHUANO", and #436 "FAMAGUSTA") because they are essentially defective or contain only one year of data (so the total stations used to form the larger set is 1,482 stations out of 1,486 total tide gauge stations).

Thus, one set of the pair has 25 stations, the other set in the pair has 1,482 stations.

III. The Software Tools

The software module that produces these graphs (by querying various SQL database

Fig. 3a

tables: WOD, PSMSL, GISS, and TEOS) uses the same process and number of years (1880-2016) to produce the data for the graph pairs.

You will notice a difference in the two a-b graph pairs, because the two pairs are based on two data (measurement) scenarios.

The graph pairs show that our world view is based on our measurement view, i.e., the measurements we take will determine what we think of the world we are inhabiting (You Are Here).

Fig. 3b

Using all PSMSL tide gauge stations, as a means of determining global sea level, can damage a sound world view with a northern hemisphere bias (which results in lower global sea level averages).

That is because tide gauges in the northern hemisphere are located such that they record an inordinate number of measurements from areas where sea level is falling  than those in the southern hemisphere do (Proof of Concept , 2, 3, 4, 5, 6, 7, 8).

IV. The Rationale

The g23 group is the result of an attempt to even out the hemispheres so as to give a more

Fig. 4a

balanced world view of sea level dynamics (The Gravity of Sea Level Change).

The careful selection process took into consideration whether the tide gauge stations are located in areas where land is subsiding or uplifting, and similar reasons.

The goal of selecting the cream of the crop locations for collecting measurements is to produce better scientific data so as to craft and form a better scientific result.

As an example, it would not matter how much data one had from a plethora of tide gauge stations located near Juneau, Alaska, combined with data from a plethora of tide gauge stations near Stockholm, Sweden.

Fig. 4b

That is because one would conclude that "sea level all around the globe is falling" just because it is falling in those areas.

Representative data from a balanced number of locations is very necessary in order to give a true picture of sea level change.

All of that reasoning is explained in the video at the bottom of this post.

V. Dates Of The Data

The subjects of the data are sea level change, atmospheric temperature change, ocean water temperature change, and ocean volume change caused by both water temperature and the addition of mass to the ocean due to cryosphere melting.

Fig. 5a

The data in my SQL database contains PSMSL (sea level) data going back to 1807, GISS (atmospheric temperature) data going back to 1880, but the WOD data only goes back to 1967.

So, I opted to begin at the year 1880, which means that I had to generate estimated WOD ocean water temperatures from 1967 back to 1880.

To do that I used, as a guide, the GISS surface temperatures (1880-2016) along with the WOD ocean temperature, salinity, depth and pressure data from 1967 through 2016 (~half a century).

Fig. 5b

I also had to generate estimated ocean mass/volume back to 1880, which began with the latest ocean volume calculations of oceanographers circa the year 2000.

I did that by paralleling the PSMSL changes (in terms of percent of decrease and increase) in sea level with decrease and increase of ocean mass (if the tide gauge recorded average sea level increased by x% or decreased by y%, I increased the mass / volume by that x% or decreased it by that y%).

Once I accomplished those calculations, I could then do the TEOS calculations as to thermal expansion (thermosteric ocean volume change).

So, as a result the graphs show both thermosteric and mass volume change.

VI. Discussion of the Graphs

Remember that the object of this post is to show that not only is measurement data important, but where the data comes from is also important (for example, if we want to know how ocean temperature has changed over time, we must measure all the way down, not just at the surface).

Let's discuss the graphs.

The graph pair at Fig. 2a and Fig. 2b show the difference between mass volume change (not thermosteric) using 1,482 stations, and for contrast also using the g23.

As you can see, there is not much difference, and the trend is identical (the measurements are the facts, but the trend is the truth).

The graph pair at Fig. 3a and Fig. 3b (g23) show the TEOS concept of conservative temperature and absolute salinity as well as a comparison of thermal expansion / contraction (thermosteric volume change) with mass volume change (ice melt-water caused volume increase).

The graph pair at Fig. 4a and Fig. 4b (g23) show the thermosteric volume change close up.

Again, you will notice that there are difference between the g23 graph and the other graph, but the measurements and calculations are the facts, the trend is the truth.

The graph pair at Fig. 5a and Fig. 5b (g23) show in situ values, including GISS surface temperature anomaly, WOD ocean temperature and salinity, as well as PSMSL sea level change since 1880.

VII. Conclusion

From now on, I am going to use the g23 data when I discuss global sea level change.

I can still use any other tide gauge station data when targeting specific locations.

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

---

---