Weekly Area of Snow Extent

Guest Post by Willis Eschenbach

I got to thinking about snow the other day. It was occasioned by my look at the correlation (both positive and negative) of temperature and albedo. Albedo is a measure of how much sunlight is reflected from the clouds and the surface. The greater the albedo, the more sunlight is reflected. Here’s the graph that set me pondering:

Figure 1. Correlation between surface temperature and albedo. Negative correlation (blue and green) means the albedo goes down (less reflected sunlight) as the surface warms. Positive correlation (red and orange) means the albedo goes up (more reflection) as the surface warms. Gray line shows zero value.

In the red and orange areas, which are mainly in the tropics, the albedo goes up as temperatures rise. This is generally because clouds form as temperatures rise, reflecting more sunlight and cooling the earth. In the blue and green areas, on the other hand, the albedo goes down as temperatures rise. Over the extratropical land, much of this change is from snow and ice. As the land warms, snow melts and the albedo goes down. And as the land cools, snow falls and the albedo goes up. This is a positive feedback, with warming leading to increased solar energy, and cooling leading to less solar energy.

One thing that is highlighted by this map is that the positive feedback from the changes in sea ice are much smaller than the feedback from the changes in snow and ice on land, for several reasons.

The first one is the small area of the sea ice variations. Note that the feedback is only in the areas that are seasonally uncovered and covered by sea ice—permanently ice-covered areas don’t have much albedo change. Net annual variation in Arctic sea ice is about ± 5 million square kilometres. This is only about 1% of the area of the globe.

Another reason the changes on land are larger is that when snow melts, it exposes soil and plants, both of which have low albedos. But when sea ice melts, it reveals ocean … and the albedo of the ocean at low sun angles is already pretty high. As a result, the melting of the ice doesn’t change the albedo as much as the melting of the snow.

Another reason the land varies more is that snow extends much closer to the equator than sea ice. As a result, the sun rises much higher over snow than sea ice, and thus the snow intercepts more sunlight than the same area of ice up near the poles.

Another reason is that as you can see from Figure 1, the negative correlation of the albedo and temperature is greater over northern lands than northern oceans.

All of this has made the snow-covered areas of the northern hemisphere the main suspects in the onset of the ice ages. The generally accepted theory is that the so-called “Milankovitch” variations in the earth’s orbit change the amount of sunshine hitting the northern hemisphere. When the northern hemisphere summer sunshine gets weak enough, the snow on the northern land doesn’t melt back as far. This residual snow reflects more sunlight, which leads to cooler temperatures, which leads to more snow, which leads to more reflected energy … I’m sure you can see the end of this story, glaciers a mile thick covering Chicago.

Now, people seem to have a strange need to believe in some kind of existential threat hanging over our heads. There appears to be a desire to worry about something, as long as it is dire and a couple of decades away. In the past we’ve filled this need by worrying about the “population bomb”, or the “ecological footprint”, or the dreaded arrival of “peak oil”. Nowadays, it seems like “global warming” is taking over the role of the scourge du jour.

Me, I prefer to only concern myself with real possibilities of real harm. We’ve seen a couple of degrees warming since the Little Ice Age, and overall the effects have been beneficial to humans, plants and animals. I have no concern about the fabled Thermageddon of a couple degrees more warming—the effects are not grave, will likely be beneficial, and I have strong doubts that it will happen this century.

Another ice age, on the other hand, seems to be both inevitable and very destructive. And to raise the stakes, near as scientists can tell the next ice age either due or overdue … this is already the longest of the “interglacials”, the historical periods in between the ice ages.

So I would suggest that we keep a fairly close watch on the snow cover of the northern hemisphere. Because when the apparently  inevitable ice age comes ’round again, it seems to me that the first sign will be an increase in the snow cover in North America and Eurasia.

Fortunately, the good folks at Rutgers University have a dataset showing the weekly area of the extent of the snow in the northern hemisphere that goes back forty years or so. Here’s that data:

Figure 2. Rutgers University snow extent data. Note the missing data prior to 1972. Data Source: Rutgers Snow Extent Data

So … how is the extent of the snow trending over time? Well, if we look at the complete data, which extends from 1972 to present, here’s how that breaks down:

Figure 3. Decomposition of Rutgers snow extent data. Top row is observations. Second row shows the trend in the 52-week mean. Third row is the regular seasonal variations. Bottom row is the residual variation once the seasonal and overall trends are removed. Note the different scales on all four rows.

The second row in Figure 3, entitled “trend”, shows the changes in the mean value over time. The snow area generally dropped during the first half of the record. Subsequently, it first rose and then remained level in the second half. So the good news is that we don’t appear to be started into an ice age. The other good news is that we also don’t seem to be headed for a time when our children won’t recognize snow … overall, like most climate records, not a whole lot going on. However, that is unlikely to last forever.

Finally, some speculation. I have long held that the main two ways that we affect local climate are through land use, and also via airborne soot (or “black carbon”) and “brown carbon”. Brown carbon is the airborne carbon from inefficient combustion of wood, coal and other fuels. In addition to coming from forest fires, brown carbon mainly comes from billions of cheap stoves and open cooking and heating fires in the developing world. Because of the prevailing winds, a goodly amount of the soot and brown carbon produced in the northern hemisphere falls on the northern snow and ice. And because the carbon compounds are dark in color, they are warmed by the sun. This leads to a more rapid melting of the snow. It has been suggested that this is the reason for the retreat of the European glaciers since the 1800s.

Now, humans have been dumping large quantities of soot into the atmosphere for quite some time now, ever since we managed to tame fire. And presumably, for all that time that soot has helped to melt the northern hemisphere snows and glaciers, so they didn’t start lingering further and further into summer. So … would it not be truly ironic if pollution, in the form of soot and brown carbon,  were all that has been holding off another ice age? And wouldn’t it be a cosmic joke if our efforts to clean up soot and brown carbon pollution were the straw that broke the back of the Holocene, and ushered in the new ice age?

Do I think that’s the case, that soot is all that is keeping the next ice age at bay? Y’know … I truly don’t have a clue whether that’s true or not. That’s one beauty of climate science, that there are so many mysteries.

I’m just saying, I’m keeping an eye on the snow extent …

w.

DATA AND CODE:

I’ve posted up a .csv file containing the Rutgers data here, and the R code to read it is here.