Subscribe to Diving Deeper

Ocean Service Feeds

What is a Podcast?

A podcast is a an audio file published on the web. The files are usually downloaded onto computers or portable listening devices such as iPods or other players.

Read more about podcasting from webcontent.gov

Find other podcasts from the US government

Diving Deeper: Episode 27 (October 7, 2010) -

Remote Sensing

(INTRO)
HOST: Welcome to Diving Deeper where we interview National Ocean Service scientists on the ocean topics and information that are important to you! I’m your host Kate Nielsen.

Today’s question is….What is remote sensing?

Remote sensing is the art and science of detecting, identifying, classifying, and analyzing features on the Earth’s surface. To do this, analysts use imagery from terrestrial, aircraft, or satellite platforms equipped with photographic and non-photographic sensors. More simply stated, remote sensing is the process of gathering information from a distance.  

To help us dive a little deeper into this question, we will talk with Chris Parrish by phone on remote sensing – what it is, why it is important, and how scientists use this data. Chris is the Lead Physical Scientist with the National Geodetic Survey’s Remote Sensing Division. Hi Chris, welcome to our show.

CHRIS PARRISH: Hi Kate, thanks for having me here today.

(BACKGROUND ON REMOTE SENSING)
HOST: Chris, is there any additional explanation that you can give us on remote sensing from this initial definition? Just to help us lay out the concept better for today’s episode.

CHRIS PARRISH: Well Kate, the definition that you gave is sort of the standard, textbook definition of remote sensing. In other words, if you look on the first page of just about any remote sensing textbook, you’ll see something like, remote sensing is the art and science of measuring things at a distance. And that definition is definitely not wrong, but I’ve always felt that it’s not all that helpful in the sense that if somebody is brand new to the field and I give them that definition, it doesn’t actually tell them that much about what somebody who’s working in remote sensing actually does or how or why they do it.

So when I explain remote sensing to people, I’ll typically put it in terms of examples of people that I actually know and work with, who work in the field. So for example, I have one colleague who works for a large commercial satellite imagery company, and that company operates Earth observation satellites, they supply imagery to the U.S. government and military as well as to other commercial firms. I also work with a lot of people, including several of my NOAA coworkers, who are involved in collecting, processing, analyzing remotely sensed data collected from aircraft so that could include aerial imagery or light detection and ranging, lidar data, that’s used for topographic mapping, and those data will be used for just a huge variety of applications. It could be anything from forestry to agriculture to military applications, ecosystem mapping and monitoring, emergency response, climate change for example. And in our office, we typically use the data for coastal mapping, particularly mapping the national shoreline.

HOST: How do we collect the data, and then once it’s collected, how is it transmitted back for analysis in the office from an aircraft or satellite?

CHRIS PARRISH: OK, well, to answer that Kate, first of all you’re definitely right that you can kind of separate things into two steps – the data acquisition and then the data processing and analysis and in between there, you’ve got to get data from point A to point B. So basically for any type of remote sensing data acquisition, there’s always some type of sensor that’s installed either in the aircraft or satellite and that’s being used to collect the data. So to use just one kind of straight forward example, I’ll start with the example of aerial imagery. So in this case, you’ve got a camera that’s mounted on an aircraft and it’s being used to collect imagery of the ground below as the plane flies over. And that camera might be in a lot of ways just like a typical camera that you could buy at your local electronics retail store, it’s probably a little more expensive, maybe more highly calibrated, maybe a little more sophisticated in terms of being able to collect data in more spectral bands, but it’s the same basic concepts. You have light reflected from the Earth’s surface, collected at the camera lens, and then used to form an image.

So the next step that you mentioned, is how do you get that data then down to the ground for processing. In the case of aerial imagery, that’s fairly straight forward. A lot of times, what might happen in the case of a digital sensor is that the imagery or other data are logged to a removable hard drive. So the plane lands, the operators pull the hard drive off the plane, and then somehow transmit that back to the office for further processing.

In the case of satellite remote sensing, it’s a little bit more complicated because you’ve got to have ground stations for the data to be transmitted down to. But in either case, once those data make it back to the office, that’s where typically the majority of the processing and analysis happens.

HOST: Thanks Chris. You mentioned sensors a few times – are there different types of sensors that collect the data?

CHRIS PARRISH: Yes, there are definitely a lot of different types of sensors. At kind of the highest level, sensor types are sometimes broken into two broad categories – passive and active. So passive sensors, what that means is that there’s an external light source that’s being used and that’s almost always the sun. So as a result of that, of course, you can only acquire data during the day. And an example of a passive sensor is an aerial camera like I was just talking about. Again, what you’ve got is reflected sunlight from the Earth’s surface that makes it back up to the camera and is collected and used to form an image.

That’s the passive sensor, the difference with an active sensor is that it uses its own source of electromagnetic radiation. So an example there is an airborne light detection and ranging, or lidar, sensor. Those are used a lot of times for topographic mapping. The idea is you’ve got a laser system in an aircraft that’s transmitting laser pulses down to the ground and using those to measure a range.

So, even after dividing sensors into those two broad types – active and passive – you can still keep going, subdividing sensors into different sub-types. The bottom line is that, yes, there are definitely a lot of different types of sensors that are used in remote sensing.

HOST: Back at the beginning we talked about all the different ways, whether it’s by plane or satellites, that we’re getting this data. How do people work into this component? What happens, I guess in the field if you’re doing data collection more that way?

CHRIS PARRISH: OK, well if we’re talking about airborne data collection, there’s typically a mission crew that’s, just to use a typical example, it might include two pilots and a sensor operator, and they travel around with the aircrafts acquiring data. We can consider an example where there’s a project over the Virginia Coast Reserve on the Virginia Barrier Islands. So, the instructions that the flight crew receives from the office maybe contain a box that’s plotted on a map showing them where the project area is that they want to acquire data for, maybe telling them what type of sensor to use, what flying height to acquire the data at, etc.

As far as a day in the life, typically what happens is the mission crew will get up in the morning and almost always the first thing they want to do is check the weather. The reason for that is that there are a lot of different types of weather that can affect airborne remote sensing. Of course, things like snow and rain, but even just having cloud cover a lot of times can prohibit data acquisition. So, if the weather looks good, they’ll maybe pick a time of day to go out and begin acquiring data and that can be based on a lot of factors including not just cloud cover, but things like sun angle, it might even include tides if you’re acquiring data in a coastal area.

And then during a flight, basically the pilots are responsible for keeping the plane flying straight and level on the specified flight lines as well as all the typical things that pilots are responsible for like communicating with air traffic control, just making sure the flight is safe, and then at the same time they’re doing that, typically the sensor operator’s in the back of the plane, they might be pressing buttons on the sensor or looking at their laptop computer that they’re using to control the flight and just make sure that all the data that they’re acquiring is good and that all the right parameters are set.

And then in the evening after a flight, usually what happens is that the person that is acquiring the data, the sensor operator, is going to do some preliminary processing. So they’re not necessarily processing it all the way through to an end product, but they’ll at least look at enough of it to make sure that the data that they acquired is good and that nothing really went wrong with the flight and then from there, they’ll transmit that data again back to the office for additional processing.

HOST: Chris, you touched on a few things like weather, tides, and sun angle that need to be considered when collecting imagery. Is there anything else? Is it a fairly routine process or is data collection really different each time and in each location?

CHRIS PARRISH: Kate, there are definitely a lot of different considerations that go into data acquisition. Usually, there’s a really extensive planning process that will start a long time before the plan will go out to a specific area to begin acquisition. And that project planning process really starts with asking some fundamental questions about how the data are going to be used.

For example, is the project intended for mapping shoreline, is the goal to evaluate habitat change, is it to map airport obstructions. And so based on the different requirements for these different application areas, the project planners can begin going in and assessing first of all things like, what type of sensor or sensors are appropriate for this project, what’s the level of accuracy that’s required in the data, and what’s the level of spatial resolution that’s needed, in other words what are the smallest features that you might need to be able to discern from the data.

And so the answers to those types of questions in turn will enable them to make decisions like what’s the right flying height for this acquisition, what are the various mission parameters that need to be set a specific way in order to ensure that they’re meeting those requirements.

HOST: Is it expensive to collect and analyze remote sensing data?

CHRIS PARRISH: Well Kate, that’s actually a really difficult question to answer. On the one hand, it can literally cost hundreds of millions or even billions of dollars to collect and process remote sensing data. And that would be the case, for example, if you’re talking about building a fleet of Earth observation satellites as well as pay load sensors, launching them into space, putting in place a network of ground stations, and then all the people that are needed to do all those different phases and process the data. As you can probably imagine, that can be just unbelievably expensive.

But then at the complete opposite extreme, it can actually be free to obtain and analyze remotely sensed data, and the reason that’s the case is that a lot of times people who have paid for a data collection, will then sometimes make those data freely available for others to use. One example of that is NOAA’s Coastal Services Center maintains a web portal called Digital Coast, where people can go and download coastal remotely sensed data, things like aerial imagery, lidar, and then use those. Actually two of my NOAA NGS colleagues, Jason Woolard and John Sellers, and I over the past two years have taught a conference workshop where we’ve showed people how to take free data, free software tools, and use them for processing and analyzing lidar data and imagery. I guess in this case though saying that it’s free is maybe a little bit deceptive in the sense that it wasn’t free to begin with, somebody obviously paid for the data collection initially, it’s just that they’ve been nice enough then to make those data available for others.

Sometimes you have things that are a little bit between those two extremes I gave. For example, if you’ve got existing satellite imagery that might not be made available for free, but the cost of acquiring them is a lot less than if you had to go out and acquire those data yourself. And then kind of another example of something in between would be the airborne data collection that we were talking about earlier. The cost there could really vary depending on the type of type of aircraft, the cost of fuel, the number of pilots, and sensor operators. But to give kind of some rule of thumb numbers, it might be somewhere between a few hundred to let’s say a few thousand dollars per hour to operate the aircraft and then maybe another couple hundred dollars per hour for any post-processing that’s done.

HOST: Chris, throughout our interview today, we’ve talked about remotely sensed data and aerial imagery. Are these the same thing or are they different?

CHRIS PARRISH: Well, aerial imagery is just one type of remotely sensed data and if you go back, Kate, to the definition that you gave of remote sensing back at the very beginning of this interview, I think you said something like, remote sensing is the art and science of gathering information at a distance, so aerial imagery definitely fits that definition, but so do a lot of other things like satellite autimetry, radar, lidar, etc.

(USING REMOTE SENSING DATA)
HOST: From our discussion today, remote sensing sounds like a rather complicated data gathering process that, in turn, gives us a lot of data and information. What are some of the real benefits of remote sensing?

CHRIS PARRISH: Kate, in answering the question of what the benefits of remote sensing are, I think it’s kind of helpful to go back to when people first started putting cameras on airplanes or even before that, going back to the 1800s when people were putting cameras on balloons or even on kites, and to ask the question, what was the motivation for doing that? And a lot of times the motivation was simply that people wanted to be able to map a relatively large area a lot more efficiently and probably a lot more cost effectively than if they had to do it through a ground survey. And I think that today, a lot of those same motivating factors often hold true.

And then just to throw in a related benefit, a lot of times, remote sensing is the only feasible way to collect data in areas that might be dangerous or in some cases, really impossible to put people on the ground to collect the data. And you can imagine a really remote area in Alaska as maybe an example of that.

HOST: Thanks Chris, great examples on some of the benefits from remote sensing data. How do we use remotely sensed data to detect oil spills?

CHRIS PARRISH: Well, one of the big benefits of using remote sensing for oil spill detection is that sometimes these oil spills can cover very large areas of the ocean and coasts. So in that case, one of the only ways to efficiently and cost effectively cover that type of large area is with remotely sensed data. An example would be, using satellite imagery to assess the extent of an oil spill. And then, within NOAA’s National Geodetic Survey, we respond when requested to fly coastal areas following a spill. So that was done, for example, following the recent oil spill in the Gulf. We mapped shoreline in Louisiana, Mississippi, Alabama using high-resolution aerial imagery, and that was done immediately after the spill, to provide a baseline data set, so that repeat observations can be made over time that will allow scientists to monitor any changes to the coast due to the spill.

HOST: How does remote sensing support emergency response perhaps following something like a hurricane?

CHRIS PARRISH: Kate, our office, NOAA NGS is pretty frequently called on to respond to hurricanes and other disasters. Typically what we’ll do is we’ll try to fly the impacted areas as soon after an event as possible. So one of the really common goals is to be able to acquire imagery and have it online for people to download within 24 hours of a hurricane making landfall and we’ll try to provide geo-referenced aerial imagery, which means that each image pixel is associated with a particular set of map coordinates, a latitude/longitude for example, and the way that’s used – first responders can use it in looking at what the most heavily impacted areas are, sometimes property owners can use it if they’ve been displaced and they want to know what’s going on with their property, was it damaged.

But then hurricanes are just one example of the types of disasters that our office is sometimes called on to respond to. Another recent example would be the Haiti earthquake this past January.

HOST: Can anyone use and apply remote sensing data or do you need someone who is specialized or trained in some way to apply this data to all of the different situations that we’ve talked about today?

CHRIS PARRISH: Kate, I noticed you phrased that question as an either/or. Can anyone use remotely sensed data or do you need highly trained people and the answer is really yes to both. On the one hand, it’s definitely true that anyone can use remotely sensed data, in fact, sometimes people use remotely sensed data, maybe without really even being aware that they’re doing so. An example of that would be, if you’ve ever used Google Earth or Google Maps to look at your house or town and satellite imagery, then you’ve used remotely sensed data.

But at the same time, it’s also definitely true that highly trained people are needed for some applications, and as an example of that, I can take a recent project that I collaborated with some folks at the Naval Research Laboratory on, and in this particular case, the goal of the project was to be able to estimate shallow bathymetry, in other words water depths, from a particular type of remotely sensed data called hyperspectral imagery and it was a pretty complex process that consisted of first trying to correct for the effects of atmosphere in the imagery and then performing mathematical calculations to be able to estimate water depth from each image pixel. So that would be just one example of a case where you really do need highly trained, knowledgeable people to be able to do a particular type of remote sensing task.

(NOAA’S ROLE IN REMOTE SENSING)
HOST: Thanks Chris, I think you’ve given us a good background on remote sensing and how the data is used. Can you expand on the role of the National Ocean Service in remote sensing? Is the National Ocean Service involved in data acquisition, or actually collecting the information, or are we more involved in the data analysis side of things?

CHRIS PARRISH: I would say all of the above. If you look across NOS, NOS scientists are involved in a whole lot of different types of projects that use remotely sensed data. This can be anything from monitoring harmful algal blooms to coral bleaching, land use changes, and bathymetry. In other big areas, is using remotely sensed data to assess and respond to threats of climate change such as sea level rise and shoreline erosion.

And as we discussed, the specific office that I work in, the National Geodetic Survey, we use remote sensing for coastal mapping. Another thing we do is that we use remote sensing technologies for conducting airport obstruction surveys in support of the Federal Aviation Administration. So really, it’s a whole host of applications and NOS scientists, again, get involved in both the data acquisition and the processing and analysis of the data.

HOST: Thanks Chris. Do you have any final closing words for our listeners today?

CHRIS PARRISH: Well Kate, in closing, I’d just like to say that remote sensing is really an exciting field to work in. It’s great to be able to provide products and services that people can use and really benefit from using remotely sensed data, so I’m really glad to have had the opportunity to share some of what we do with the listeners today.

HOST: Thanks Chris for joining us on Diving Deeper and talking more about remote sensing, how data are collected, and, most importantly, how we use this data. To learn more, please visit geodesy.noaa.gov.

(OUTRO)
That’s all for today’s show. Please tune in for Diving Deeper Shorts in two weeks.

(top)