Friday, January 2, 2015

Know your data - ocean acidification again

There's been another article (archived here) on the World Ocean Database (WOD) pH data, which was previously the subject of a wrong and malicious article alleging fraud. (Levitis et al (2013) provides an introduction and background to the database.)

It's kept some of the deniers at WUWT busy examining the data - and some who aren't deniers, too. While others just transferred via keyboard whatever random thought popped into their head.

My apologies in advance. This is another too long article - I got carried away playing with a new toy I found. Click read more if you want.


Willis Eschenbach wrote this third article. He says he doesn't believe the scientists committed fraud. The scientists being Richard Feely and Christopher Sabine. What Willis wrote was:
So … was there “pHraud” in not utilizing this data? I say no, there was no fraud. I say this in part because it’s so difficult to infer intent. Because I have been falsely accused of having bad intent a number of times, I’m sensitive on the subject. I dislike accusations without evidence, and I see no evidence of fraud in this case.

He also said that the Mike Wallace's chart below is "wrong". I've added a notation and highlights to Mike's chart to show up some of the more obvious problems:

Dumb chart from Mike Wallace with HW annotation.
Source: WUWT

The outliers make no sense. In addition, there are entire years missing. But they aren't the only problems, as I describe further down.

Throwing non-factual data about at WUWT


Willis also said that he found it bizarre that someone would throw out data, writing:
I was unimpressed by the graph in that post, which seemed simplistic and, well, in a word, wrong. But on the other hand, I certainly found it bizarre and most interesting that someone would throw out that huge amount of scientific data. That was the reason I forwarded it to Anthony, in the hope of unraveling the actual truth of the matter. 

There's no-one throwing out data - but Willis is certainly throwing false allegations around. I don't know why he thinks that anyone threw out data. In fact, Sabine et al (2004), discusses how scientists have spent several years compiling records from myriad sources:
Recognizing the need to constrain the oceanic uptake, transport, and storage of anthropogenic CO2 for the anthropocene and to provide a baseline for future estimates of oceanic CO2 uptake, two international ocean research programs, the World Ocean Circulation Experiment (WOCE) and the Joint Global Ocean Flux Study (JGOFS), jointly conducted a comprehensive survey of inorganic carbon distributions in the global ocean in the 1990s (4). After completion of the U.S. field program in 1998, a 5-year effort was begun to compile and rigorously quality-control the U.S. and international data sets, including a few pre-WOCE data sets in regions that were data limited (5). The final data set consists of 9618 hydrographic stations collected on 95 cruises, which represents the most accurate and comprehensive view of the global ocean inorganic carbon distribution available (6). As individual basins were completed, the ocean tracer–based ΔC* method (7) was used to separate the anthropogenic CO2 component from the measured DIC concentrations (8–10). Here we synthesize the individual ocean estimates to provide an ocean data-constrained global estimate of the cumulative oceanic sink for anthropogenic CO2 for the period from ∼1800 to 1994 (11).

Another conundrum - is Mike Wallace a bit slow, or is he a climate disinformer?


Think on that - five years. And Mike Wallace has the cheek to claim, based on his own inept "analysis" and blind ignorance:
The world's ocean pH measurements have been omitted from the modern narrative of ocean acidification, as documented at www.abeqas.com under the category 'ocean ph and acidification'.
In particular, the true records have been replaced by a misrepresentation of recorded ocean pH in a presentation to the U.S. Congress (the "feel2899.pdf" reference used in Markey U.S. Congressional 2011 hearings on climate change and ocean acidification (House Report 111 – 709)).

He wrote that in a petition he's set up. What an utter nutter. Is he really that dumb or is he faking it and pandering to deniers? Remember, not only has he pored over the data, he reportedly exchanged emails over a period of several months with some of the world's leading experts on ocean acidification. One would presume he's read at least some science on the subject as well. Given all that, it's hard to believe that an ageing PhD candidate one-time hydrologist could be that dense (though in my experience, from people I've met, high intelligence is not mandatory to become a PhD candidate). So it's a toss up between him indulging in some disinformation for deniers and him being too thick to know what he's talking about.


More data - but it's sparse


There is data going back to 1910 in the WOD, but it's patchy. There is even an ocean acidification portal, which suggests inclusion of other parameters for pH calculation. It's freely available to anyone who wants to use it. You don't have to stick to pH measures, you can calculate it. There is also heaps more data elsewhere, such as the Argo floats (data here, if you're game); the coral reef watch program data. I expect there is a lot more, too.

Willis lumped all the data points together in a couple of maps. I don't know what he thought he was achieving, but all he showed was that over the past few decades, people have been taking pH measurements in various different parts of the ocean.


Collecting and interpreting the data


Nick Stokes turned up some interesting stats on the WOD pH data. I did some investigation of my own. I found the World Ocean Database software, which is totally wonderful and not that hard to learn. I generated some plots and statistics from it.

First let's look at the ten year period from January 1911 to 31 December 1920. There are only 446 data points in the entire ten year period - and that's global, with the only constraint being the data is limited to the top 200 m. The mean pH is 8.18. But the pH range (not shown in the chart below) is 0.62 to 8.32. So that tells you a bit about the quality. The chart below just shows pH vs depth for the pH range from 6 to 10. The red line simply shows one of the stations to depth (it starts at a higher pH and goes lower as it gets deeper):

Data source: NOAA World Ocean Database

And here is where the measurements were taken each year - as an animated gif, showing the spots for each of the ten years from January 1911 to December 1920. Click to enlarge it.

Data source: NOAA World Ocean Database


Now go back and look at Mike Wallace's chart and tell me what was going through his head when he included data from 1910 to 1920. Not just included data, but reported it as "global ocean pH". There was no data at all for seven of the ten years! Of the remaining years, there were separate cruises in four separate parts of the ocean where data was collected. None at all in the southern hemisphere. None at all in the Pacific Ocean. None at all in the Indian or southern oceans. None of the four cruises can reasonably be compared with any other.


Some points worth noting


Someone made the point that global pH is not as meaningful as local pH. Most studies of ocean pH do seem to be around a particular locality. For example, there is a lot of concern about how a drop in pH is affecting coral reefs, which are important as marine sanctuaries, biodiversity and neighbouring fisheries. There have also been studies estimating global changes over time.

An important point is that pH doesn't change all over the ocean at the same rate. This is important when trying to get an estimate of global change. You can't just lump all the data together and get an average. From the IPCC WG1 AR5 section 3.8.2 report (my emphasis).
A global mean decrease in surface water pH of 0.08 from 1765 to 1994 was calculated based on the inventory of anthropogenic CO2 (Sabine et al., 2004), with the largest reduction (–0.10) in the northern North Atlantic and the smallest reduction (–0.05) in the subtropical South Pacific. These regional variations in the size of the pH decrease are consistent with the generally lower buffer capacities of the high latitude oceans compared to lower latitudes (Egleston et al., 2010).

What that means is that if you have a whole heap of data from the North Atlantic, and just a few data points from the sub-tropical South Pacific, and toss all that data into a single heap and calculate the average, then you'll think that globally pH has dropped more than it actually has. You can't do that and get a sensible result.

Finally, you don't need direct pH measurements to estimate ocean pH. You can use dissolved inorganic carbon (DIC). This is described in the supporting online material from Sabine et al (2004).

More data problems


Even though the data is sparse for ocean pH prior to the establishment of pH monitoring in the late 1980s, I was curious to see if there was a shift in the mean pH. (I wonder how much of the data post late 1980s is not in the WOD - for example, from satellites - plus data derived from other measurements. WOD only includes direct pH measures from cruises.) I tried the period from 1951 to 1960 but there were too many measurements that were weird. Not only are some of the measurements the pretty well impossible, but there is a huge conglomerate of in the top 10 metres or so that defy comprehension.

Here's a scatter plot so you can see what I mean. What the animation shows is all the data points from January 1951 to December 1960, morphing into those of one particular cruise for the same period. Click to enlarge it.

Data source: NOAA World Ocean Database

You can see that one cruise accounts for a quite a bit of the strange data (very low pH), though not all of it. It will be other cruises that have the odd data in the first few metres.


A non-scientific comparison - pre-1960 and the past 10 years


The next thing would be to select an area for comparison. One that didn't have strange and incomprehensible readings. So I chose part of the Pacific Ocean for no particular reason.

First I have to stress that this is in no way pretending to be a scientific analysis. For one thing, the cruises weren't overlapping. I haven't limited to any particular seasons. The data is sparse. Also, as I pointed out earlier, you can't mix and match data from the low latitudes with that from the high. At least not without doing some area weighting - which I didn't do. And there are a heap of other reasons I'm sure you can think of. To show what I mean, here are maps of the data points. Click to enlarge:

Data source: NOAA World Ocean Database

You can also see that the recent period is more in the southern hemisphere than the earlier period. Both have data from high and low latitudes but I haven't worked out the number of data points in each. Which is why you shouldn't take the results as having much meaning.

Anyway, here is a comparison of the aggregate pH data. First the pH histograms where you can see a distinct shift down in pH:

Data source: NOAA World Ocean Database

When you look at the data in more detail, you can see more differences, such as the slightly wider pH range in the earlier data; the difference in the number of data points (see the map above as well); a difference in the mean depth of readings; and the difference in the range - with the recent data reporting nothing shallower than 1.8 m.


Jan 1931-Dec 1960
Jan 2005-Dec 2014
pH Mean
8.11
7.93
Std Dev
+- 0.17
+- 0.16
pH Range
7.5 - 8.4
7.4 - 8.2
No. Points
1351
3150
Mean depth (m)
75.7
84.8
Depth range (m)
0 - 200
1.8 - 200


The bottom line


Mike Wallace is either abysmally ignorant or practicing to become a climate disinformer.

You really need to know what you are doing and be clear about what you are trying to achieve if you want to get any meaning from climate data. I suggest a lot of reading beforehand. Sabine04 (listed below) is not a bad place to start. Chapter 3 of the IPCC report has more information and heaps of references for further reading. There are a couple of other references in my previous article on the subject too.

Based on the best evidence science has to offer, ocean pH is dropping. And that makes sense, as I said at the beginning of my earlier article.



From the WUWT comments


It was quite fun reading some of the WUWT comments. There were a number of people who had a shot at analysing the WOD from a file that a WUWT reader called Simon Filiatrault provided, together with a helpful document describing what he did. Most WUWT-ers who manipulated the data didn't understand what they were doing, apart from being able to play with data. What I mean is, they didn't understand the data itself. Being able to play with numbers and data sets isn't sufficient.


Willis Eschenbach and his 2.5 million observations
December 30, 2014 at 9:01 pm
I told Anthony about the story, not because I thought it indicated fraud, but because I found it quite bizarre that a pile of data that big, 2.5 million observations made by professional oceanographers over decades, would just be ignored.
Note that the accusation of fraud came from Marita Noon, not from me. As I said above, I’m reluctant to ascribe motives. Heck, half the time I’m not sure what my own motives are until after the fact, and perhaps not even then.
w.

His 2.5 million observations are from 53 cruises and 259,209 stations. A station is a single location, each of which can have up to around 20 data points, each at different depths. The cruise with the most data points (1,341,611) and most stations (126,949) had data from 29 Jul 1924 - 18 Sep 2008 and covered the world from east to west and 73.1ーS to 89.6ーN.  (The number of data points drops to 1,579,913 different observations if you only count in the top 200 m of ocean, which is where acidification has most effect for us and marine life. )

As an aside, the pH data go from pH=0 to pH=24,041, with 214 data points above pH 14. A quality filter helps.


Andrew is one of the many WUWT hopefuls who really, really wants to think the oceans are getting more alkaline - he can't have any chemistry or physics background.
December 30, 2014 at 8:35 pm
Yes well. Readers shouldn’t HAVE to trawl through data to construct time series where prima facie an interesting scientific truth emerged from recent analysis showing a rising pH.
We spend literally $trillions. And if ocean acidification is disproven, that may be one of the most valuable discoveries in history.

old44 thinks not having cruise vessels churn through the Great Barrier Reef means no-one cares.
December 30, 2014 at 9:09 pm
Strange, with all the talk of ocean acidity destroying coral reefs, no samples taken on the lower half of The Great Barrier Reef where all the fertiliser from agriculture and where the coal exporting ports are located, and very few from Mackay upwards, the same for Ningaloo Reef. it is almost like they were trying to hide something.

Which only goes to show that the WOD data is not the only and perhaps not the main source of information scientists use when doing research on acidification at different locations. I've added a link to a paper about the GBR acidification.  Perhaps someone else can add more insight on that score.


ferdberple really, really wants Mike Wallace to be "right" and proceeds to lump various data from disparate regions and disparate seasons together in a time series trying to prove it so.
December 31, 2014 at 7:39 am (extract)
My conclusion looking at the data is that the explanation that it is not of sufficient quality to calculate a pH trend is without foundation:...

That's enough. Most of the WUWT rabble use the article as an excuse to shriek "fraud, fraud" and more. They haven't a clue about science. Most of them don't even know the names of the scientists they are defaming, but they know all the denier catch phrases.




Schlitzer, R., Ocean Data View, http://odv.awi.de, 2013 (the software)

Levitus, Sydney, J. I. Antonov, O. K. Baranova, T. P. Boyer, C. L. Coleman, H. E. Garcia, A. I. Grodsky et al. "The World Ocean Database." Data Science Journal 12, no. 0 (2013): WDS229-WDS234. http://dx.doi.org/10.2481/dsj.WDS-041 (open access)

Sabine, Christopher L., Richard A. Feely, Nicolas Gruber, Robert M. Key, Kitack Lee, John L. Bullister, Rik Wanninkhof et al. "The oceanic sink for anthropogenic CO2." science 305, no. 5682 (2004): 367-371. DOI: 10.1126/science.1097403 (pdf here)

Wei, Gangjian, Malcolm T. McCulloch, Graham Mortimer, Wengfeng Deng, and Luhua Xie. "Evidence for ocean acidification in the Great Barrier Reef of Australia." Geochimica et cosmochimica acta 73, no. 8 (2009): 2332-2346. doi:10.1016/j.gca.2009.02.009 (pdf here)

11 comments:

  1. This 2012 OA Report Card for Australia gives a good overview of recent research with a "bonus" 12 minute video interview. The report is chock full of references as well.

    Another report is due this year, but I can't see where the funding will come from as all the organisations behind the last report are either disbanded or "refocused". I have regular contact with a number of the authors so they can give me the latest news after the Christmas break.

    ReplyDelete
    Replies
    1. Thanks, Anthony. I'd not seen that site before. It looks great. Keep us in the loop.

      Delete
  2. There should be a flag in WOD indicating that the pH reading of 0.62 should not be trusted. The rules for flagging are conservative - only the obviously wrong are flagged.

    ReplyDelete
    Replies
    1. Yes, quantpalaeo, there are lots of flags in WOD, not just for outliers. In this exercise I was just playing with the data so I didn't bother downloading the flags. (I've since done so). I'm thinking the WOD might be going to distract me a bit this year. There is a huge amount of data. And having found such a great software package that can handle very large amounts of data beautifully (without having to teach myself R or some other application) is very enticing. The biggest hurdle is getting the data. pH on its own was fine, but some of the data requests generate enormous downloads, as you'd imagine.

      I also wanted to show how WUWT seemed blissfully unaware of any quality issues with the data. Anthony wrote something about "no caveats" showing he needs spoon-feeding, and can't be bothered reading the documentation or understanding the options for downloading.

      Most of the others at WUWT - well outliers were the least of their problems with the way they mishandled the data.

      Delete
  3. This comment has been removed by a blog administrator.

    ReplyDelete
  4. Sou, can you please point to your scientific CV? It's unprofessional to not disclose your credentials when leveling scientific assertions as if you were an expert. I hope you will understand if I conclude that you are not being candid or honest with your readers (all four of them).

    I am Mike Wallace, the researcher whom you have featured in this post, and my credentials are proudly posted at my website [redacted direct link in line with the comment policy - archived web page is here. Sou.]. I would like to reach out to all four readers of Sou's post and encourage them to explore the additional information on ocean pH data misrepresentations I've posted at [redacted direct link in line with the comment policy - archived web page is here. Sou.]

    There will be more such posts and papers to come, by others in addition to me, until even Sou will be embarrassed for his/her advocacy of ocean pH data fabrication and omission.

    [I just noticed this comment from Mike Wallace. It was originally posted April 27, 2015 at 3:02 AM. I had to edit it to comply with the comment policy, so the date reflects that. Sou.]

    ReplyDelete
    Replies
    1. Seems Mike Wallace finds navigating a website as difficult as understanding ocean acidification.

      Delete
    2. Mike Wallace - the "science" you're offering up is so poor that anyone that takes the time to examine it can only laugh - as Sou has easily shown. One doesn't have to look much further than your treatment of the data between 1910 and 1920 to see that you haven't a clue how to use the data in a scientific manner. Your mistreatment of the WOD ph data has been pointed out to you before. Apparently you don't learn - just parrot your mistakes.

      Delete
    3. "There will be more such posts and papers to come, by others in addition to me, until even Sou will be embarrassed for his/her advocacy of ocean pH data fabrication and omission."

      Please promise us that you'll continue to post your thesis here, with referenced data and analysis. You can be assured that your work will be well scrutinised.

      [I haven't chuckled so much since the Orssengo W- (or was it 'M-'...?) curve.]

      Delete
  5. If anyone's interested, I just saw a reference to Mike Wallace pop up again over here in our local paper. The article is about drilling in ANWR.

    ReplyDelete

Instead of commenting as "Anonymous", please comment using "Name/URL" and your name, initials or pseudonym or whatever. You can leave the "URL" box blank. This isn't mandatory. You can also sign in using your Google ID, Wordpress ID etc as indicated. NOTE: Some Wordpress users are having trouble signing in. If that's you, try signing in using Name/URL. Details here.

Click here to read the HotWhopper comment policy.