Wednesday, July 29, 2009

Noctilucent Emissions

There's a new study suggesting that "noctilucent clouds" are caused by the plumes of space shuttle launches, which is somewhat ironic for me. For those that don't know, noctilucent clouds are the highest clouds found in the Earth's atmosphere. They are ice crystals floating high in the mesosphere, where large amounts of moisture are not typically found, and are usually only seen in twilight when the setting sun illuminates them against the dark sky. There is also evidence that these clouds are a modern phenomenon, within the past century, and so they may be related to human activity, or perhaps climate change.

Noctilucent clouds over Lake Saimaa. Photograph taken by Mika Yrjölä. Permission by Creative Commons Attribution 2.0 License (from Wikipedia).

Early studies have suggested that noctilucent clouds were caused by space shuttle launches. The space shuttle exhaust plume is composed mostly of water vapor. As the shuttle launches into orbit, it can dump significant amounts of water vapor into the upper atmosphere as it passes through it. More recent studies by Dr. Michael Kelley have added credence to that idea: shuttle launches in 2003 and 2007 produced corresponding noctilucent clouds.

NASA launched a satellite called AIM in 2007 to study noctilucent clouds. AIM stands for Aeronomy of Ice in the Mesosphere. AIM continues to study the phenomenon using several instruments, including two imagers and a meteoric dust measuring device.

The part I find ironic is that NASA launched a satellite to study a phenomenon caused... by the launch of NASA satellites!

OK, there are far more noctilucent clouds than can be entirely explained by shuttle launches, so this is not a complete exercise in navel-gazing. The shuttle-noctilucent connection was known before AIM was selected by NASA.

Update 2009-07-30: Added noctilucent cloud image from Wikipedia.

Wednesday, July 8, 2009

Solar Dud or Doozy?

There is some news in the past few days that the a few new sunspots are appearing, it may be that the next period of solar activity has begun. Will it be a dud or a doozy?

The sun has a quite regular cycle that repeats about every 11.5 years. Each cycle represents a reversal of the the sun's magnetic field -- magnetic north becomes south, south becomes north -- so it actually takes 23 years for the sun's magnetic field structure to return to its starting configuration.

During one of these 11.5 year cycles, the sun's magnetic field gets tangled and wound up in its circulating convective zone, located in the outer third of the sun. Once in a while, the magnetic field pokes out from the beneath the surface, and a sunspot appears. So sunspots are indicators of the how chaotic the magnetic field is within the sun.

Solar activity is a problem for us on the earth since the energetic particles ejected during solar storms can affect communications, power grids and the orbits of satellites. Being able to predict the the solar cycle, both timing and strength, is a valuable tool that can save lives and equipment. Unfortunately, predictions have been more of an art than a science.

I was really intrigued by a presentation at the 2005 AAS conference several years ago by Peter Gilman and Mausumi Dikpati, which claimed an improved method for solar cycle prediction. Their solar model showed that the solar convective zone had large scale circulations, almost like oceanic currents on earth. They also showed that it takes approximately three solar cycles for the solar flows -- which are almost like conveyor belts -- to make one circuit. Thus, they could train a reasonably accurate predictive model, based on the known solar activity from three cycles before (with the added benefit, that it produces predictions for about three cycles into the future as well). Their prediction was that the current solar maximum would be delayed by 6-12 months, but 30-50% more intense than the previous cycle. The "conventional" predictions were calling for the beginning of the cycle to begin in early 2007, while Dikpati and Gilman's group were calling for activity starting in the late 2007 to 2008 time frame.

Well at this stage, it's clear that both the conventional and new Dikpati/Gilman predictions were wrong, since we're past halfway into 2009 before any serious solar activity has appeared. But it's interesting that the onset of the cycle has indeed been delayed from its expected appearance, which indicates that perhaps there is something behind the Dikpati/Gilman model. The "conventional" prediction was recently revised, and now claims that the next cycle will be a dud -- weaker than usual. I've had a little harder time determining if the Dikpati/Gilman group has revised their forecast for the strength of the cycle.

Either way, I think it will be an interesting cycle to watch. I actually hope this solar cycle is an extreme -- either a dud or a doozy -- rather than an average one. Extremes are much better test cases for theories than boring average cases. While I have something to lose if this cycle is a strong one, at least it would be lost in advancement of science.