What do you know about the polar bears apart from the fact that they live up north in the Santa Land. 
How many of these white furry mammals do you think are there alive on this planet? 1,000? 10,000? 100,000 or may be more? What do you think is happening to these numbers over the years? Are they going to become extinct over the coming years? Are the numbers increasing or decreasing? If they are decreasing, what is the main factor that drives that decrease in numbers?

Well to answer some, if not all, of these questions, we decided to analyze a visualization of the Polar Bear Population Data.
The original visualization can be found at : http://pbsg.npolar.no/en/dynamic/app/

click the image to launch the original visualization

There are an estimated about 25,000 polar bears left in the world, and the population is expected to decline over the coming years.
The data can be viewed by sub-population, by nation, or by eco-regions.

Visualization by Sub-Populations

The first visualization shows a nice arctic view of the earth divided according to the 19 bear sub-populations. Each of them color coded according to the trend observed in the population number.

The legend is shown on the top left corner which makes it easier to understand the color coding of the map.

This visualization gives a good overview of the trend of the polar bear population according to each sub-population. Some a declining trend, some stable population and some increasing. Here deficient data is also taken into account. The regions with deficient data are not just overlooked but indicated the deficiency of the data in an efficient way.

Mouse over gives a nice sneak peek of the information about the data.

The color selection is poor. Green is a poor color choice since green/red is the most common form of color blindness. So definitely, a person with some form of color blindness will have a hard time in interpreting the distinct regions on the map. 

 Protanope SimulationDeuteranope SimulationTritanope Simulation
         Deuteranope Simulation                                                Protanope Simulation                                            Tritanope Simulation

You can click on geographic regions to see more details about the region which include a text overview and time series for more troubling numbers on polar bear killed by humans and pollution.

The time series gives a nice overview of the data about the harvesting of polar bears. Nice old school 2-D bar chart with no reflections makes it very easy and convenient to interpret the graph and the data. The time graph also shows the annual harvest and the trend in polar bear harvesting which related to the graph and helps in understanding the plot better.

The pollutant levels show the mean pollutant levels(NG/G) for the four compunds BDE-153, HBSD, CHL, Mirex for the female, male polar bear and the cubs as well in the form of text. 
Using some form of visuals to represent the numbers for mean pollutant levels would have been better as the human tends not to pay much attention to the text and numbers when visualizing some data set. And besides some visuals might give a better insight of what the data is about and what the user wants to interpret.

When you click on the time series or the pollutant levels you will see the data at a higher granularity.

Annual Harvest

The use of a simple bar graph to show the data for each sub-population for all years and highlighting the area for the selected sub-population is probably a good idea but looking at the graph, its difficult to make out which bar corresponds to which year unless one mouse overs to  one of the years.

There are two types of plots one can see, one is a plot for each sub-population, gives the user to compare with other sub-population. Kind of a good idea, but the bars mixes with each other for a particular sub-population.

Another plot is where one can compare the selected sub-population with the global data for each year. 

The color coding is kept the same throughout according to the four known trends (increasing, declining, stable and data insufficient).

Pollution Studies

The graph for environmental pollutants data is another bar graph, with three lines per sub-population per environmental pollutant. 
This graph lacks a good legend as its difficult for the user to make out what the dots stand for unless the user mouse overs to one of the bars.
Its a good idea to plot the graph for all the three types, female, male and cub, on the same graph as this would allow the user to compare all the three at the same time.

Color coding according to the family (female, male and cub) could have been a good idea.

Visualization by Nations

The map is now divided according to the nations and bear population is shown for each nation with relative bear size. 
Keeping the map in blue scale is a good idea. The regions can be selected and viewed separately.

 On mouse over, we can see how many bear populations the nation has out of the total 19. This also gives us an idea about how well the nations are doing in preserving the species.

When you click on a nation, that will give you more information about the nation in the form of text and will also tell you what all sub-populations exist in that regions. If analyzed carefully one can know the common sub-populations among the nations.

The sub-populations are still color coded according to the trend of increase or decrease in population (increasing, declining, stable and data deficient) as was done in the first visualization.

This visualization also shows a circle plot, but due to the lack of a good legend/information, its hard to interpret what the circles represent. 

Do the circles represent the bear population or the number of bear sub-populations a nation has?
What do the overlapping of the circles represent? All these questions remain unanswered. Had there been some way to get the information, the visualization would have been more effective.

Visualization by Ecoregion

In the last visualization, the arctic region is divided into ecoregions. Each ecoregion giving details about the sub-populations and estimated population on mouse over.
There is not much interactivity with this view. Moreover the distinction between the ecoregions is not very effective. 

Color coding would have proven much effective or using other distinctive and prominent patterns would have more help.

About the Data

The data for this visualization consists of three data sets: one for the population of polar bears, second about the polar bear harvesting, land last one about the environmental pollutants. 

The polar bear population survey is conducted independently by organizations in each nation/sub-population 1-2 times or even less every decade. The data for some regions/nations (data deficient) is not available because they have not been surveyed by their nation or was too old.

The lack of data makes the visualization only 50% effective.

The visualization shows only two causes for the decrease in the population of the polar bears. Other factors might be the major contributing factors which were not taken into account which made the estimate of the population less accurate.
So incorporating that data might help in accurately identifying the trend in the polar bear population.

What can one ask and find?

Polar bears are at the top most level in their food chain. So what's causing the decrease in their population?
You might want to know how many polar bears are still left and how has the trend been over the past years?

Well for that, the division of the map according to the sub-population provides a good overview about the population estimate and the observed/calculated trend of the population.
Hovering the mouse over the selected region will open a pop up that provides all the relevant information.

If you want to ask, what is the effect of global warming and increasing temperatures on the polar bears?

Then this is one of those things that the application can't do. This is due to the lack of data relating to the global warming. Had it been included, would have provided a much broader aspect to the reason as to why the polar bear population is decreasing.

One of the other questions that you might want to ask is that "who is having more effect of the environmental pollutants? the females, males or the cubs?"

The visualization provides an efficient way of providing you with the information you need.
The environmental pollution graph provides a nice and easy visualization of the same. Each sub-population is divided into three graph plots, one for female, one for male and the third for the cub.
Also when you mouse over the plot, you can see the table/text version of the same which you can  co-relate with the plot.

Which sub-population has more strict laws when it comes to polar bear harvesting? Or which all sub-populations are working towards saving the polar bears from being extinct in their region due to excessive poaching?

The annual harvest plot provides a thorough comparison of each sub-population for each year. So not only you can compare each sub-population with each other but you can compare the corresponding values for a sub-population for every year. This gives a good idea about the trend in the harvesting of the polar bears. 

After viewing the first visualization (by sub-population), one question that arises in the mind would be "How come the M'Clintock Channel has an increasing trend in the population as compared to the decreasing trend of the rest of the sub-populations?

This can be explained effectively by looking at the annual harvesting and environmental pollution plots of the M'Clintock Channel and comparing it to the rest of the sub-populations.

One can see that the average annual kill in M'Clintock Channel sub-population is only 2 polar bears. May be the laws are strict there. May be there were deterrence programs that were more successful in this sub-population.

Another reason might be the environment of M'Clintock Channel. This sub-population happens to have a very less or no environmental pollution. As explained by the researchers "Polar bears are exposed to high levels of pollutants because of a process called biomagnification. Many pollutants are lipophillic that is they are deposited in the fat of the animals that consume them. Because animals in the arctic marine ecosystem are highly dependent on fat for storing energy, growth, insulation and buoyancy, they eat large quantities of fat causing these pollutants to rapidly accumulate in animals higher up the food chain.