Thursday, December 14, 2017

Oceans: Abstract Values vs. Measured Values - 6

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
This is a quick update concerning the previous post discussing WOD measurements possibly being outside the maximum / minimum parameters of Appendix 11 in the WOD User Manual (Oceans: Abstract Values vs. Measured Values - 5).

The post showed thermosteric sea level changes that were out side those boundaries, so I began to look into the matter.

The graphs posted show that WOD temperature measurements are not the problem.

All of the various in situ measurements, TEOS Conservative Temperature (CT) calculations, and Abstract maximum / minimums calculations show that the WOD measurements are within bounds.

It remains to be seen what the problem with the thermosteric sea level changes that were out of bounds is caused by.

I am still researching the issue to determine if it is a TEOS or Dredd Blog bug concerning thermosteric volume calculations.

I will update this series accordingly when I know more.

I moved things around in Fig. 1 so you can see the colors of the Abstract Maximum / Minimum lines (the long ones from 1880 - 2016).

The WOD in situ and CT lines are short red and black lines spanning 1968 - 2016 (right hand side of the graphs).

The Abstract maximum and minimum abstract CT lines are greenish, while the middle (average) abstract CT line is blue.

As you can see, the various configurations of layers and zones are reasonably within the confines of the maximum / minimum ranges.

The salinity measurements were out of bounds (higher than the maximum) so I am checking that out prior to posting the Absolute Salinity graphs.

I am checking the source code closely to see if there is a logical error that would make the salinity go out of bounds.

Even though I doubt that a two or three g/kg value above the maximum, which is what the salinity is showing, would cause the sea level gyrations in the previous post, I am looking closely at all of that.

I suspect that it is more likely to be a TEOS issue, because both the thermosteric volume and the thermosteric sea level gyrations are related.

The previous post in this series is here.

Monday, December 11, 2017

Oceans: Abstract Values vs. Measured Values - 5

Fig. 1 Abstract maximum, average, and minimum
Just in case the message is not getting through, today I present some more graphs to show how true the three scientific papers (which I quoted from yesterday) were  (Oceans: Abstract Values vs. Measured Values - 4).

I mean where those papers pointed out how the measurements which scientists have been able to take are not spread out evenly (in terms of latitude & longitude) across the vast oceans of the world.

My argument or discussion about this is that we need to have a way of doing with WOD datasets what the GISTEMP and PSMSL data users have been able to do with those datasets.
Fig. 2a
Fig. 2b
Fig. 2c

That is, to define which WOD layers and/or zones can be used to represent the entirety of the ocean conditions, that is, the oceans as a whole.

Only twenty-three PSMSL tide gauge stations out of about 1,400 tide gauge stations can do that in terms of sea level change.

The GISTEMP is similar in that the global mean average temperature anomaly can be shown in the same manner (a representative subset).

But, as those papers discussing the world oceans point out, the same is not yet accomplished with WOD datasets.

The measurements are too concentrated in certain areas to the exclusion of other areas, are too few, and do not go deep enough into the abyss.

The ARGO automated system of submarine drones is changing that in the upper 2,000 m of the oceans, but that is a relatively recent technological win.

There is no long term in situ set of measurements of ocean temperature and salinity data going back in time for over a century, like there are with the PSMSL and GISTEMP datasets.

To visually point out the measurement aberrations I am speaking of, let's look at the new graphs generated by version 1.7 of the software I am constructing.

The graph at Fig. 1 shows the ABSTRACT (calculated) maximum, average, and minimum thermal expansion and contraction pattern from the years 1880 to 2016.

The three graphs at Fig. 2a - Fig. 2c show what happens when in situ WOD measurements are added to the data stream used to generate those abstract graphs.

Fig. 3 Abstract avg. compared to measured
The patterns made by the in situ measurements are out of sync with the abstract patterns made with the WOD information about valid maximum and minimum temperature and salinity values.

Since those values from the WOD manual define validity at all ocean basins and all depths at those basins, being out of sync with them is a problem, especially when the out-of-sync pattern emerges using any of the three different sets of WOD data which compose three different layer lists.

Those three sets are 1) all layers, 2) 6 selected layers, and 3) 8 selected layers, as shown by the report below.

The software module loading sequence proceeds from 1 through 6 (GISS data loader, ABSTRACT data generator, G6 loader, PSMSL loader, G8 loader, and the WOD all-layers loader.

Those modules load in situ measurement data from SQL tables, as well as WOD maximum / minimum valid values.

The software then organizes the data into annual structures (past to present).

The measured values are converted into TEOS values, according to the TEOS rules, by functions in the TEOS toolkit (e.g. Golden 23 Zones Meet TEOS-10).

I may have to stop using WOD layers, to instead use individually selected WOD Zones.

I want to find locations that stay within the guard rails of the valid WOD maximum / minimum values in Appendix 11 of their user manual (see links here).

I am on the case.

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

A printout of the loading sequence of the module follows:

DREDD BLOG
GISS, PSMSL, WOD & TEOS
Data Analyzer Report
(ver. 1.7)
=======================

(1) GISS Loader
---------------
processed 137 rows


(2) ABSTRACT Calculator

-----------------------
processed:
137 years of data
30 ocean basins
at 33 depths


(3) WOD G6 Loader
-----------------
processing layer 5
processed 118 rows
(59 years) of data

processing layer 7
processed 102 rows
(51 years) of data

processing layer 8
processed 162 rows
(81 years) of data

processing layer 9
processed 176 rows
(88 years) of data

processing layer 10
processed 172 rows
(86 years) of data

processing layer 12
processed 142 rows
(71 years) of data


(4) PSMSL Loader
----------------
processed 10,199 rows


(5) WOD G8 ALT Loader
---------------------
processing layer 3
processed 142 rows
(71 years) of data

processing layer 5
processed 118 rows
(59 years) of data

processing layer 7
processed 102 rows
(51 years) of data

processing layer 8
processed 162 rows
(81 years) of data

processing layer 9
processed 176 rows
(88 years) of data

processing layer 10
processed 172 rows
(86 years) of data

processing layer 12
processed 142 rows
(71 years) of data

processing layer 14
processed 142 rows
(71 years) of data


(6) WOD Loader (all layers)
---------------------------
processing layer 0
processed 60 rows
(30 years) of data

processing layer 1
processed 118 rows
(59 years) of data

processing layer 2
processed 96 rows
(48 years) of data

processing layer 3
processed 142 rows
(71 years) of data

processing layer 4
processed 188 rows
(94 years) of data

processing layer 5
processed 118 rows
(59 years) of data

processing layer 6
processed 178 rows
(89 years) of data

processing layer 7
processed 102 rows
(51 years) of data

processing layer 8
processed 162 rows
(81 years) of data

processing layer 9
processed 176 rows
(88 years) of data

processing layer 10
processed 172 rows
(86 years) of data

processing layer 11
processed 142 rows
(71 years) of data

processing layer 12
processed 142 rows
(71 years) of data

processing layer 13
processed 144 rows
(72 years) of data

processing layer 14
processed 142 rows
(71 years) of data

processing layer 15
processed 156 rows
(78 years) of data

processing layer 16
processed 102 rows
(51 years) of data

processing layer 17
processed 0 rows
(0 years) of data

Friday, December 8, 2017

Oceans: Abstract Values vs. Measured Values - 4

Fig. 1a
Fig. 1b
I. Background

This series is about what to do about the dearth of in situ measurements of the whole ocean, top to bottom (Oceans: Abstract Values vs. Measured Values, 2, 3).

The issue, including what to do about it, has been addressed in the scientific literature:
"Prior to 2004, observations of the upper ocean were predominantly confined to the Northern Hemisphere and concentrated along major shipping routes; the Southern Hemisphere is particularly poorly observed. In this century, the advent of the Argo array of autonomous profiling floats ... has significantly increased ocean sampling to achieve near-global coverage for the first time over the upper 1800 m since about 2005. The lack of historical data coverage requires a gap-filling (or mapping) strategy to infill the data gaps in order to estimate the global integral of OHC."
(Ocean Science 2016, Cheng et alia, emphasis added; PDF here). Going back a bit further, the issue came up in another paper:
"A compilation of paleoceanographic data and a coupled atmosphere-ocean climate model were used to examine global ocean surface temperatures of the Last Interglacial (LIG) period, and to produce the first quantitative estimate of the role that ocean thermal expansion likely played in driving sea level rise above present day during the LIG. Our analysis of the paleoclimatic data suggests a peak LIG global sea surface temperature (SST) warming of 0.7 ± 0.6°C compared to the late Holocene. Our LIG climate model simulation suggests a slight cooling of global average SST relative to preindustrial conditions (ΔSST = −0.4°C), with a reduction in atmospheric water vapor in the Southern Hemisphere driven by a northward shift of the Intertropical Convergence Zone, and substantially reduced seasonality in the Southern Hemisphere. Taken together, the model and paleoceanographic data imply a minimal contribution of ocean thermal expansion to LIG sea level rise above present day. Uncertainty remains, but it seems unlikely that thermosteric sea level rise exceeded 0.4 ± 0.3 m during the LIG. This constraint, along with estimates of the sea level contributions from the Greenland Ice Sheet, glaciers and ice caps, implies that 4.1 to 5.8 m of sea level rise during the Last Interglacial period was derived from the Antarctic Ice Sheet. These results reemphasize the concern that both the Antarctic and Greenland Ice Sheets may be more sensitive to temperature than widely thought."
(The role of ocean thermal expansion, AGU, emphasis added). Basically, the scientists point out that this exercise is not a picnic:
"The oceans present myriad challenges for adequate monitoring. To take the ocean’s temperature, it is necessary to use enough sensors at enough locations and at sufficient depths to track changes throughout the entire ocean. It is essential to have measurements that go back many years and that will continue into the future.
...
Since 2006, the Argo program of autonomous profiling floats has provided near-global coverage of the upper 2,000 meters of the ocean over all seasons [Riser et al., 2016]. In addition, climate scientists have been able to quantify the ocean temperature changes back to 1960 on the basis of the much sparser historical instrument record [Cheng et al., 2017]."
(The Most Powerful Evidence, Inside Climate News, emphasis added). That quote contains a reference to "Cheng et al. 2017" which contains the following statement:
"In this paper, we extend and improve a recently proposed mapping strategy (CZ16) to provide a complete gridded temperature field for 0- to 2000-m depths from 1960 to 2015.
...
The success of a mapping method can be judged by how accurately it reconstructs the full ocean temperature domain. When the global ocean is divided into a monthly 1°-by-1° grid, the monthly data coverage is [less than]10% before 1960, [less than]20% from 1960 to 2003, and [less than]30% from 2004 to 2015 (see Materials and Methods for data information and Fig. 1)."
(Improved estimates of ocean heat content from 1960 to 2015, Cheng et al, 2017). In other words, it has not yet been accomplished.

That is where the Dredd Blog criticism of "thermal expansion is the major cause of sea level rise in the past century or so" comes from (On Thermal Expansion & Thermal Contraction, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27).

II. An Abstract Observation

Now, let's consider that "gap-filling (or mapping) strategy" exercise (mentioned in the last quote above, Cheng et al, 2017).

The problem has been approached here at Dredd Blog by developing a software program I call an abstract pattern generator. which produces WOD patterns using data in the WOD documentation.
Fig. 2a
Fig. 2b
What I mean by "WOD patterns" can be seen at Fig. 1a and Fig. 1b.

The upper pane of Fig. 1a is a graph generated from data of the Permanent Service for Mean Sea Level (PSMSL).

It details sea level rise (SLR) at the "Golden 23" tide gauge stations (185.157 mm of SLR).

The lower pane is a graph of the abstract pattern of thermal expansion over the same time frame (25.019 mm of SLR).

In other words, the abstract thermal expansion pattern shows that thermal expansion is only 13.5% of total SLR (25.019 ÷ 185.157 = 0.135123166 or 13.5%) during that time frame, which means that it is not a major portion of global sea level rise, because those numbers also mean that 86.5% of SLR is caused by ice sheet and land glacier melt water.

The abstraction calculation is based on World Ocean Database (WOD) data in their official documentation, depicted in part at Fig. 2a and Fig. 2b, which is a portion of "APPENDIX 11. ACCEPTABLE RANGES OF OBSERVED VARIABLES AS A FUNCTION OF DEPTH, BY BASIN" (see Appendix 11, page 132, of The WOD Manual, PDF).

The gist of Appendix 11 is to show maximum and minimum values at all ocean depths in all ocean basins around the globe.

By adding the maximum and minimum values together, then dividing by 2 (at each depth of each ocean basins), the software is then ready for the next step, which is to conform those values to GISTEMP constraints.

By "GISTEMP constraints" I mean adjusting those mean average Appendix 11 values by the GISTEMP anomaly pattern.

That is done by multiplying the WOD values by 0.93 (93% of that GISTEMP anomaly value becomes the temperature anomaly value in each ocean basin at each depth).

That is because scientists tell us that some 93% of heat trapped by green house gases ends up in the oceans.

So, by fusing that GISTEMP anomaly pattern to the abstract temperature pattern made by the WOD data, we have a pattern which we can use to generate Thermodynamic Equation Of Seawater (TEOS) patterns (e.g. Golden 23 Zones Meet TEOS-10).

III. Using Abstract Patterns With TEOS-10

Using abstract WOD data to generate TEOS values is done by the same process as using in situ ocean temperature and salinity measurements (The Art of Making Thermal Expansion Graphs).

To conform either in situ temperature and salinity measurements or abstract temperature and salinity values, one uses the TEOS functions (the difference is that the abstract values have been conformed to the GISTEMP pattern as stated in Section II above).

The graph at Fig. 1b shows the resulting TEOS Conservative Temperature (CT) and Absolute Salinity (SA) patterns that emerge on an annual basis from 1880 - 2016 when one uses this technique.

From that, we then can then generate the thermal expansion coefficient and the thermosteric volume change.

From that thermosteric volume change we can calculate the sea level change (SLC) as shown in Fig. 1a.

Using the WOD manual data for all 30 ocean basins around the globe, and all 33 depths in each of those ocean basins, forms a pattern against which we can judge the general completeness and general accuracy of our in situ measurements.

It also helps us to select a "Golden 23" group of areas that mirror the whole ocean  (On Thermal Expansion & Thermal Contraction - 28).

IV. Comparing In Situ Measurement Patterns
With Abstract Calculated Patterns

So now we can talk about the current techniques of using what is described as skimpy in situ measurements (down to only about 2,000 m depth, when the average ocean basin depth is 3,682.2 m ... ~50% not used) to do the estimations all of the science team authors wrote about.

They pointed out that we have to use estimations in any case, because the datasets are incomplete in various places for various reasons, from dangerous conditions to weaker technology in times past.

To me, incomplete data is a bad place to start, having realized that the in situ measurements, although quite accurate and plentiful, are a patchwork of convenience-based expeditions that can make it difficult to see the entire picture.

I mean the total picture which must be constructed from outside the convenience zone of only expeditions to safe and warm global ocean areas.

That is why I hypothesize that it is better to start with an abstract pattern which matches the pattern made by our historically complete datasets (e.g. GISTEMP & PSMSL).

V. Conclusion

"He say one and one and one is three ... come together ..."



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