how to read a satellite image

As the Escaype founders, our goal isn’t just to get you the most accurate weather forecasts possible for photographers – it’s to show you how all of this stuff works, how we do what we do.

The ability to read satellite images is one of the essential skills in sunset forecasting, and a couple of tips can help give you an edge in seeking out the best light for your shots. No big science words without explanation, I promise.

There are two main types of satellite images we use in the office: visible and infrared.



Let’s start with the simplest one: a visible satellite image is just a picture taken by a satellite flying over us. Thin high clouds will appear dim, because they’re translucent. Very thin, high wispies may hardly be visible at all on this satellite. Denser low clouds, however, will appear bright white.

Not too bad, right? Think about it like this: if the clouds are thin enough that you can see the sun through them (i.e. they’re translucent), a satellite looking down from above should be able to see the Earth through those same clouds. But with thick clouds, or fog, you can’t see the sun through them – so we shouldn’t expect a satellite camera to be able to see the Earth’s surface through them.

Visible satellite images, thus, are best used to identify low and mid-level clouds, as they tend to be denser.

One caveat about visible satellite images: for obvious reasons, they’re useless at night. 

This image shows a small fog bank on the San Mateo Coast, lots of clouds in the Sierra, and a fog bank in Monterey/Carmel, and lots more fog out to sea. 

Sure enough, Carmel is fogged in. 



An infrared (IR) satellite image measures temperature. How is this useful to us, and why do clouds show up at all, then?

Well, first, let’s ask ourselves a question. When you go to a very high elevation (like in the mountains), what happens to the temperature? Is it warmer or colder than at sea level?

If you said colder -- you’re correct. In most cases, in the layer of the atmosphere where weather happens, the temperature drops as you go higher.

This goes for clouds, too: higher clouds are colder than lower clouds. (In fact, most mid and high clouds are made of tiny ice crystals, while low clouds may be water droplets)

So, a question for you: why is fog barely visible on infrared satellite images, if at all? Why is it so hard to tell apart from the Earth’s surface?

(Hint: what kind of temperature difference do you expect between very low clouds and the ground?)

Thus, high clouds on infrared satellites will appear as bright white, mid-level clouds will be medium grey (a medium-grey solid mass is often “popcorn clouds”), and low clouds are dim grey, and fog might barely be visible at all.

Based off this IR satellite, from approximately the same time as the visible satellite shown above, we expect to see lots of mid/low clouds over the Sierra (our model calls them mids, but remember, the mountains are already high elevation, so the clouds look lower), lots of low clouds in Oregon, and some low clouds and fog out to sea but very difficult to make out detail. A large mass of high clouds is approaching the Oregon Coast. In Utah and Colorado, you can see clouds of varying heights... we'll get to that soon. 

This brings us to the biggest weakness of satellite imagery: it only shows you what's on top. If there are multiple cloud decks, and the top cloud deck is not translucent, you won't be able to tell what's below the upper deck. This is why we often need reports for fog and low clouds -- when there is full mid-level cloud cover, it may be impossible to see what's underneath without having reports from the ground. In some cases, there is no magic way to know from above, and there is simply no substitute to real-time reports. That's why we work together as a team around here. Of course, the same goes for mid and high clouds: you might see bright colors on the satellite in Utah, but there could be a bunch of mid-level clouds below them. 

We also note that fog is bright white on visible satellites (because it's dense), and is barely visible on IR satellites (you figured out why, right?), while for high clouds the opposite is true.

Lastly, as a guideline, clouds are usually moving, and usually from west to east (we have another post below on why that is). You can track cloud movement for the last several hours on a satellite animation, and use those trends to base your expectations for the next hour or two. (You just made a forecast!) If you look closely, you'll often see clouds forming and dissipating, not just moving. This can make prediction difficult, but fortunately our models tell us when the conditions are ripe for clouds to form, which allows us to forecast it. As for simple cloud movement (meteorologists call it advection), there's really no magic and it's nothing you couldn't do on your own. It's just time-consuming and can be an act of precision; making an estimation just one frame off on the satellites could be the difference between getting skunked and catching an epic burn. While there are some cases with very complex cloud setups that are best left for the big boys to analyze, there are a lot of times that you'll be able to see our forecast for a 100/0 sunrise, check the satellite at 5am, and you'll be able to tell yourself, yup, this is happening. And there will be times when you'll check the satellite at 5am during a sunrise watch, see the big blue hole heading for SF, and you'll go south instead and score. 

Of course, we're happy to do all the hard work for you. It's part of our service. But when you take that trip outside our official forecast area, and we can't wake up at 5am to check in with you on that 100/0, if you can take a quick look at a satellite image and say yep, it's about to blow up, you're in great shape to maximizing your success with Escaype. 

(Note: Some IR satellites use bright colors in addition to white. Don’t let this variance trip you up, bright colors and white are high cloud tops. It’s the same data, just presented with colors.)


So, now we know if there are clouds, and what kind, which is super important. How do you know the orientation is favorable for a burn? More on that later, but if you want a hint...


Satellites discussed above: 

Infrared (IR)


Animated IR

P.S. I often get asked about radar. Photogs all over the country have messaged me, saying they checked the radar and see some cloud cover. This is a misconception: radar detects precipitation, not clouds. Of course, if wet stuff is falling from the sky, it’s probably cloudy, but if wet stuff isn’t falling from the sky, we can’t say it’s clear. We use radar most commonly for detecting precipitation, which may indicate that a location is socked in, as well as tracking localized storms for lightning, though that is rare in the bay area.

Text and original site content copyright © 2016 Escaype. View our Terms of Service and Privacy Policy