Questions about the Historical Global Mean Sea Level Graphs from the University of Colorado
Posted by pwl on April 25, 2011
Looking back in time at the Mean Sea Level graphs from the University of Colorado I noticed some things that bother me about their graphs and the manner of their presentation. Some serious questions where raised.
I made this slow 2 second blink comparison movie (it speeds up at the end) of two graphs from sealevel.colorado.edu to visually compare a noticeable change in plotting and data from 20041119 and about a month later on 20041223. The graph format changed and possibly the data points where changed or deleted.
HD 720p and 1080p resolution is available.
The 60 day smoothing doesn’t show the peaks in the later graph that is showed in the earlier graph. This is a statistical distortion and is disturbing.
I find this practice deceptive as the smoothing line hides the actual data and thereby distorts the graph giving a different impression; they should plot the data with a line between the data points and then add the smoothing line.
Mean Sea Level Graph from 20041119:
Mean Sea Level Graph from 20041223:
The earlier graph shows more red data points at the 1997-1998 peak, a cluster of red dots that are not shown in the later graph a month later. This is of concern, what happened to those data points? Where they deleted resulting in data distortion fabrication? What happened to them? Are they just not plotted? If so why not?
There are also many other red dot data points that seem to move from their values or vanish completely in the later graph, and vice versa! That is disturbing and hints at data manipulations. They better have a good explanation that isn’t just hand waving.
Looking back in time at the sealevel.colorado.edu main page the graphic they present changes as they add more data on the right side… however they seem to stop plotting peak data points on a number of the graphs. I don’t know if that is just their plotting program settings or if it is an attempt of some form of manipulation, it is curious though. A detailed analysis should be done to see if the graphs have other hints of data manipulations of earlier data in later graphs. If they do they better have damned good and already well documented reasons for such mannipulations.
Also they don’t expand the size of the bit map at all, you think as they added data they’d make the bit map wider but they don’t. The effect is that the slope of the sea level rise steepens. I find that this practice in climate science to be deceptive. In presenting information to humans you must keep the graph consistent in scale. Altering the x-axis scale as you add data will lead to misconceptions about the data. I’m surprised that scientists don’t know this, or maybe they do and take advantage of it. Clearly they should provide a version with the same scale if they wish to be honest and have scientific integrity.
One lesson for the climate scientists is that when they make their presentations of graphs they need to always provide a link to the data used in the graph they are showing preferably with the commands to the graphing software to reproduce the graph exactly pixel for pixel. They also need to show the RAW unmodified data as part of their chain of custody to prove it’s not been modified or distorted by their “smoothing” function or other statistical games. In other words they need to show their work on public web pages just as they are supposed to do in scientific papers. They seem to fail to realize that they now have caught the attention of a serious scientific literate audience in the public who won’t put up with their shoddy work and graphs and lack of access to the data they represent in their communciations. They also need to explain why they change the graph when they change it. Sure as some have suggested they need to learn to communicate better, they need to follow basic protocol of the scientific method and communicate ALL the steps of their work in visual and numeric forms! They need to back up their conclusions with the data and graphs step by step. In other words they need to communicate their work by showing it!