Reading a Weather Forecast for Astronomy: Seeing, Transparency and Cloud

Every new observer learns to check the cloud forecast. Rather fewer learn that a cloudless night can still be a bad night, and that the two ways it can be bad are completely different from each other and call for completely different responses.

The vocabulary is worth getting straight, because astronomy forecasts use these words precisely and general weather forecasts do not use them at all.

Transparency: how much light gets through

Transparency describes how cleanly the atmosphere passes light. Dust, humidity, smoke and high thin cloud all reduce it. When transparency is poor, faint objects get fainter, the sky background brightens because there is more material to scatter light pollution back at you, and your images lose contrast in a way no amount of stacking recovers.

Poor transparency is the enemy of faint, broadband targets: galaxies above all. If transparency is bad and you had planned a galaxy, change the plan.

The counter-intuitive part: some of the most beautifully clear-looking nights, the ones where the air feels crisp after a cold front, can have mediocre transparency because of high-altitude moisture invisible from the ground.

Seeing: how steady the air is

Seeing describes atmospheric turbulence, the churning of air cells at different temperatures that makes stars twinkle. Good seeing means a star image holds still and sharp; bad seeing means it boils.

Seeing limits fine detail. It is the dominant constraint on planetary imaging and on splitting close double stars, and a much smaller factor on large, diffuse nebulae, which have no fine detail to smear in the first place.

This is the single most useful practical distinction in this article: bad seeing is survivable for deep-sky work; bad transparency is not. A turbulent but clear night is a perfectly good night to image the Rosette. A hazy but steady night is not a good night to image M33.^[1]

Cloud forecasts lie in a specific way

General forecasts report cloud cover as a percentage over a wide area and a long window. For imaging, what matters is whether cloud crosses your particular patch of sky during your particular two hours, and a "20% cloud" night can mean either a completely clear evening or a procession of cumulus that ruins every third frame.

Satellite imagery in motion is more useful than a percentage. So is the distinction between cloud types: high thin cirrus is the transparency killer that most often gets forecast as "clear".

Humidity, dew point and the end of your session

The gap between air temperature and dew point predicts when your optics will fog. A narrow gap means dew early, and once the corrector plate fogs your session is over unless you have a heater.

This is worth checking specifically, because it is the difference between planning a three-hour session and planning a ninety-minute one. If the forecast has the temperature falling to within two degrees of the dew point by midnight, either bring the dew heater or plan to finish early.

A practical routine

1. Check cloud first, using an animated satellite view rather than a single number.
2. Check transparency second. If it is poor, pick a bright target and lower your ambitions rather than cancelling.
3. Check seeing third, and only really care about it if you were planning planets or a small target.
4. Check the dew point gap to decide how long the session can realistically run.
5. Check the moon, which is the one variable you can predict weeks ahead and the one most often forgotten.

The habit worth building is to decide the target after reading the forecast, not before. Most frustrating nights are not caused by bad weather; they are caused by pointing at a galaxy on a hazy night because the galaxy was the plan.

Notes & sources

1. Definitions of seeing and transparency and their differing effects on planetary versus deep-sky observing. Sky & Telescope, observing basics ↩