Bathy Lidar: Harder Than It Looks


The majority of people working with lidar elevation data use airborne topographic lidar or perhaps mobile lidar. However, in the coastal zone, there is a high demand for nearshore bathymetric lidar, especially for shallow areas that can’t be cost effectively covered by survey ships. Bathymetric lidar has been around for many years now (not a new technology), so why does bathymetric lidar cost 5 to 10 times more than topographic? Today we’re going to look at the difference between topography and bathymetry from an airborne lidar sensor.

A small sample of bathy lidar in Lake Superior. Elevations are in the NAVD88 datum.
A small sample of bathy lidar in Lake Superior. Elevations are in the NAVD88 datum.
Seeing Through the Water

The first problem is water penetration.The majority of topographic lidar sensors use a wavelength in the infrared, typically 1064 nm in the U.S. and 1550 nm in Europe. Those wavelengths will only go a few centimeters in water before they’ve lost most of their power. You need to use a wavelength that isn’t absorbed rapidly by water. If you had pure water, that would be somewhere around 440 nm. Ocean water isn’t pure, especially in the coastal zone, and absorption by chlorophyll in the blue tends to push the wavelength of maximum penetration into the green region. If you frequency double a 1064 nm laser, you get 532 nm (green) output at the cost of some power.

The Need for Power

So, now you’re using a wavelength where you have a chance, but you’re still limited. The green wavelengths may be where you can get maximum water penetration, but that’s not the same as air. You’ll often see references to how many Secchi depths a system can penetrate. A Secchi depth is simply an approximation of water clarity. It is based on a broader spectrum than our laser in combination with the human eye to determine the depth at which a disk disappears from view. A better value might be the attenuation coefficient (k) at 532 nm. This describes the exponential decay of light with depth according to the equation Ed = Eoe-kz . In rough terms, this means that for every 1/k meters you go down, you’ve lost ⅔ of the laser light you had. Since our laser light has to go both down and up, you can double the number of meters traversed in figuring out your power loss.

Ok, so you’re losing a lot of power, just pump up the output on the laser, right? Not so fast. We’ve got a couple of problems here. The first is simply how much power you can pump out of a laser. You’re only going to get so much power out before you start to fry your lasing medium. On top of that, you have to divide up that power across your pulses, so higher power equals lower pulse rate. The second is eye safety. There are regulations to follow to make sure you don’t blind someone that happens to look up at your low flying plane. You’ll spread your laser beam to deal with that and probably end up illuminating something on the order of a square meter on the water surface. In that case, trying to get multiple points per meter is a bit silly.

Diagram showing the interactions for a bathymetric lidar laser pulse. Source – USACE.

Diagram showing the interactions for a bathymetric lidar laser pulse.
Diagram showing the interactions for a bathymetric lidar laser pulse. Source – USACE.
Refraction

If I put enough power through the system at an appropriate wavelength to get to the bottom and back, I should be good, right? Just calculate the range, all the angles to figure out where I was pointing, etc., and I should have my answer. Physics says “no.” For topography, the things in the way, like trees, just block some portion of the light and reduce your signal. Similarly, in the water, the absorption reduces your signal. Water does something else though. It bends the light and slows it down. It’s not a straightforward calculation anymore. We have to know where it bent, how much it bent, and how much it slowed down. If we can figure out where the water surface is, this isn’t too bad. After all, we know the index of refraction to get the speed of light in water. We know the angle at which the laser pulse hits a flat water surface (let’s not think about waves right now). If we have the water surface, we’ll know where it bent and can figure out the rest. Unfortunately, a lot of the time with just a green laser, you can’t tell where the water surface is. Many bathy systems will use an additional infrared pulse to determine the water surface. That’s yet more complexity in the system and the processing. The system also can’t use simple response thresholds to determine object returns the way a topographic lidar can. You’ll get response from scattering within the water column and due to that exponential signal decay, the response near the top of the water column is likely larger than the response from the bottom surface. This means you have to do waveform analysis.

Simply More Complex

So, what does this all boil down to? Bathymetric lidar systems are far more complex than topographic, they fly low and slow to get the power they need, and the processing is much more complex. All of these add up to systems that cost a lot more to run. On top of that, water clarity is a huge issue. If you can’t get enough signal to the bottom and back, you can’t see it. You can use a lot of time and money sitting on the ground waiting for clearer water. This can be worse than waiting for the clouds to clear in the tropics (been there, done that), especially in the surf zone. All told, it makes bathymetric lidar expensive relative to topographic lidar. Keep in mind though, it’s still cheaper than trying to get a ship in many places and sometimes a lot safer too.

10 comments

  1. I agree with everything you said and appreciate the straightforward way it was presented.

    There is one problem that deserves technical attention, however, and that is the surface return or splash, which usually overloads the detector to the point that it is useless. The key question is for how long? In building a foliage penetrating lidar, similar problems occur. The first leaf that is struck produces huge returns that cause most detectors to be recovering during the time in which maximum sensitivity is needed. And I appreciate the eye-safety problem as a new laser safety officer would not let my beam-spread eye-safe laser be used because the laser INSIDE the lidar was not eye safe.

    However, there is a technical solution for the detector at least.. Hamamatsu, whom I have no financial interest in promoting, has created a Hybrid photodetector that can recover to photon counting sensitivity of 45% QE in one nanosecond after complete overloads. This is not a PMT or Pin diode with an amplifier which is sometimes called a hybrid tube. Rather, it is a photocathode with 8500V acceleration into a non-geiger APD which further amplifies the single photo-electron to usable photon counting levels. It is possible to count photons at 1 GHz or more, and with a 400 ps rise time, the resolution is excellent.

    There is one more technical point that should be considered for any serious researcher. That is the background radiation from the sun which overloads the detector due to the wide bandpass of the receiver filters. A one Angstrom filter at 532nm will allow reasonable backgrounds on detectors with reasonable receiver sizes in daylight. However, these tend to be expensive – about $6000. Even one nanometer filters allow too much light into the detector unless the field of view and collection optics are reduced to unacceptable parameters. A ten nanometer filter probably would only be useful at night.

    One more point for those how love technical discussions: Most PMT amplifers cannot count photons or even be linear beyond about 10 MHz. This is because the maximum allowable average anode current is exceed. And Geiger mode APDs are useless for Bathymetry because the recovery time tends to be nearly a microsecond after surface splash.

    In summary, a 4″ to 6″ receiver with a wide field of view may be used with a hybrid detector to count photons at 1 to 5 GHz with sub-nanosecond bins in a 1 Angstrom system for Bathymetry. Then a reasonable choice of laser may be made to penetrate to desired depths using photon counting which allows accumulation of lidar returns not possible with analog signals.

    My interest is that I have been making GHz count rate photon counters and amplifier-discriminators for many years as well as lidars for atmospheric probing.

    Like

    • So going by your details above, bathy LIDAR is quite complex at least in an oceanographic scenario. What about in a littoral/’brown water’ scenario (ie attempting to map the bottom of a fair sized river). Do the same constraints apply, or are they different?
      I am considering attempting to supply de-formable data for the purposes of predicting likely outcome in an erosive issue in a moderate sized river.
      Any assistance would be appreciated.

      regards

      Rob Frith

      Like

      • Rob,
        The essential physics are the same in a river system. Simplifying greatly, the more stuff in the water, the more difficult it is to see the bottom. So, the dynamics that are suspending particles or growing algae are going to affect your ability to detect the bottom.

        Kirk

        Like

  2. So going by your details above, bathy LIDAR is quite complex at least in an oceanographic scenario. What about in a littoral/’brown water’ scenario (ie attempting to map the bottom of a fair sized river). Do the same constraints apply, or are they different?
    I am considering attempting to supply de-formable data for the purposes of predicting likely outcome in an erosive issue in a moderate sized river.
    Any assistance would be appreciated.

    regards

    Rob Frith

    Like

    • Hi Kirk,

      Thanks for your reply (and apologies for the double post).

      So Bathy LIDAR might not work in a river with a large amount of suspended silt/particulate matter (say after heavy run-off on steep hill country). That’s a pity…was thinking I might have found the Silver Bullet that I could use to advance my project…

      What about actual speed of the water in terms of speed over ground. By this I mean, if the river is traveling at 4 knots/2.1 metres per second like many of our rivers do here in New Zealand, would this affect the accuracy of the measuring pulse? I mean, obviously the water within the ‘column’ through which the LIDAR pulse travels is not necessarily going to be the same water at the end of the measurement, as it was at the start?

      Is LIDAR capable of picking up invasive water-based vegetation, at least in terms in indicating excessive build-up? Such build-ups might indicate potential erosive or blocking actions caused by that vegetation, and/or the medium or base it is growing in and around (say a fallen tree beneath the surface, not visible normally but the sheltered downstream ‘shadow’ of that obstruction)…

      Hope I’m not outwearing my welcome, your assistance in this is greatly appreciated.

      Regards

      Rob

      Like

      • Rob,
        Right, a large amount of suspended silt or particulate matter is going to greatly reduce the depth of penetration. We’re all still looking for that silver bullet.
        About the speed of the water, it really makes no difference. While light in water isn’t quite as fast as light in air, it’s still pretty darn fast. The light is going to be in the water column for well under a microsecond, so your water has moved less than the width of a hair. Plus, I don’t think it’s speed is going to matter regardless. Think of the case where you’re looking at a tree. If the wind starts to blow, it doesn’t change your perception of how far away the tree is.
        The lidar should be able to pick up vegetation, assuming it is sufficiently reflective. Depending on how much vegetation you’ve got, it may be that you only get the vegetation and don’t get to the bottom, so you may have some work to do to figure out what the pulses reflected off of. I would expect vegetation is going to be a much rougher looking surface than the true bottom.

        Hope that helps,
        Kirk

        Like

  3. Hi,
    I have to give a detailed report about airborne bathymetric lidar in college , so i was going in deep details about it. Can you please explain in details about the behaviour of laser pulse as it enters water since its behaviour is different from that in air?

    Like

    • Apoorva,
      The basic physics of light transmission through a media don’t change, it’s just the properties of the media that are changing. There are lots of places for a college student to look those up, so I won’t go into too much detail here. You do have to account for the change in direction at the boundary between two media due to the index of refraction change (just as you would for any optical analysis). Within the media you have absorption and scattering of photons. Both are much higher in water than in air and the wavelength dependencies are different. The absorption and scattering are also increased by other things in the water (phytoplankton, suspended sediments, colored dissolved organic matter, and so forth). All of this makes bathy lidar more complex, but the primary impacts are reductions in the maximum detectable depth and spreading of the laser footprint on the bottom.

      Like

      • Thanks for the quick reply and I forgot to mention your article is the simplest and most nicely put one to explain bathymetric lidar. I would be thankful if you can give a little description(of labels) on the image you used since I am not able to completely understand what is going on in it when the beam enters water..
        Regard,
        Apoorva

        Like

  4. Apoorva,
    I’m not sure exactly what part you’re having trouble with. Perhaps it is that the diagram doesn’t have the refraction at the surface explicitly labeled. At the air/water interface, the photons in the incident beam with either reflect or enter the water after undergoing refraction according to Snell’s law. That’s all that’s going on there.

    Kirk

    Like

Leave a Reply. Comments are moderated.

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s