Shop OBEX P1 Docs P2 Docs Learn Events
Measuring snow depth — Parallax Forums

Measuring snow depth

The snow will be flying up here in New England in the not to distant future and this year I want to try building a snow depth sensor. I have found a commercial sensor based on ultrasound for around $150 and a few internet posts using the relatively inexpensive Parallax Ping sensor.

Has anyone tried using the Parallax 28044 laser range finder to measure snow depth? This sensor has a 2 meter range which would work quite well for measuring normal snow fall amounts around here. However, snow being what it is I wonder if the laser would reflect consistently off the surface so it could be measured accurately, say +/- 1-2 mm. At $100 it is a bit too pricey to buy just to see if it might work.
«1

Comments

  • Will freshly fallen snow reflect a Ping burst for accurate measurement? How many locations will you monitor, since prevailing winds at the time of snowfall tends to scoop it out of one place and pile it higher in another.
    A really low tech solution would be to just bungy some cheap yardsticks to several trees or other fixed objects and observe with a pair of binoculars and average the readings.
  • True, snow being fluffy will not be a good acoustic reflector. However there are commercial sensors out there based on the accoustic pulse echo method for measuring distance, albeit starting around $150 and up

    I have a nice copper snow depth gauge that I normally use but it's hard to see when I'm away from home :-))
  • I have a nice copper snow depth gauge that I normally use but it's hard to see when I'm away from home :-))

    Wireless security camera linked to WiFi?
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2015-11-11 17:14
    With a wifi video link to a security camera, you could simply use one of these....good for about 5 feet of depth.

    http://www.walmart.com/ip/Disney-Peel-Stick-Growth-Chart-Winnie-the-Pooh/15676326?action=product_interest&action_type=title&item_id=15676326&p

    Of course, making your own would be even cheaper. It seems everyone uses a stick which can be as much as 60 feet long. http://www.tripadvisor.it/LocationPhotoDirectLink-g143020-d144747-i43765188-Crater_Lake-Crater_Lake_National_Park_Oregon.html

    https://en.wikipedia.org/wiki/Snow_gauge
  • Around these parts of the woods there are naturally-occurring (slightly) radioactive sands. (I think it's from the mass wasting of thorium-bearing strata further up the mountain.) The idea is to mount a pole in the ground using concrete made from this sand. Then atop the pole put a Geiger-Müller tube. This way you are measuring the actual water content of the snow and not just its depth.
  • BeanBean Posts: 8,129
    I think the 28044 sensor would work fine. It doesn't do "time of flight" measurement. It simply looks for where the laser dot it appears in the camera.
    Snow should show up very well. I would be willing to bet that it would work quite well.

    I have one I'd be willing to let you borrow if you want to try it. Send me a PM if you are interested.

    Bean
  • With a wifi video link to a security camera, you could simply use one of these....good for about 5 feet of depth.

    http://www.walmart.com/ip/Disney-Peel-Stick-Growth-Chart-Winnie-the-Pooh/15676326?action=product_interest&action_type=title&item_id=15676326&p

    Of course, making your own would be even cheaper. It seems everyone uses a stick which can be as much as 60 feet long. http://www.tripadvisor.it/LocationPhotoDirectLink-g143020-d144747-i43765188-Crater_Lake-Crater_Lake_National_Park_Oregon.html

    https://en.wikipedia.org/wiki/Snow_gauge

    I'm liking the Pooh gauge idea and a web cam!!
  • Bean wrote: »
    I think the 28044 sensor would work fine. It doesn't do "time of flight" measurement. It simply looks for where the laser dot it appears in the camera.
    Snow should show up very well. I would be willing to bet that it would work quite well.

    I have one I'd be willing to let you borrow if you want to try it. Send me a PM if you are interested.

    Bean

    Got any snow where you are yet?

    I was thinking about making some type of "snow simulator" then trying the Ping sensor to see how it might respond. Perhaps something like a fluffy blanket lightly stretched over some type of wooden frame might mimic the soft acoustic properties of snow. If the Ping works then I have a really inexpensive sensor.

    But using a laser would be way cooler! I may take you up on your kind offer once we have snow on the ground to test it with. Alternatively, I could try to behave over the next several weeks in hopes Santa would leave a 28044 under the tree this year ;-)
  • It seems everyone uses a stick which can be as much as 60 feet long

    If I needed A 60 foot stick to measure snow depth I think I might be looking for some other real estate!
  • MJ,
    You've probably already run across it, but just in case, have you looked at Chris Nafis' site, http://howmuchsnow.com/? He has prototypes based on both ultrasonic and infrared triangulation, along with proof data and background.

    Here is an old thread from the BASIC Stamp forum from when Chris started work on the ultrasonic method.

  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2015-11-12 02:42
    MJHanagan wrote: »
    With a wifi video link to a security camera, you could simply use one of these....good for about 5 feet of depth.

    http://www.walmart.com/ip/Disney-Peel-Stick-Growth-Chart-Winnie-the-Pooh/15676326?action=product_interest&action_type=title&item_id=15676326&p

    Of course, making your own would be even cheaper. It seems everyone uses a stick which can be as much as 60 feet long. http://www.tripadvisor.it/LocationPhotoDirectLink-g143020-d144747-i43765188-Crater_Lake-Crater_Lake_National_Park_Oregon.html

    https://en.wikipedia.org/wiki/Snow_gauge

    I'm liking the Pooh gauge idea and a web cam!!


    That's a relief... I thought you might be offended.
    Personally, I rather have a typhoon any day than a winter in New England.
  • Personally, I rather have a typhoon any day than a winter in New England.

    But there is a reason you don't see typhoon scenery featured on many holiday greeting cards! :lol:
    A good coating of the white stuff covers up some otherwise bleak brown colored scenery.
  • Sorry, I just spent too many early mornings scraping ice from the windshield, and then driving to work in a car that only got warm and cozy as I was arriving at work.

    When a typhoon hits, we get a day off. The other days are T-shirt weather and travel by motor scooter.

    Digging out the car and the driveway just adds to one's toil.
  • Sorry, I just spent too many early mornings scraping ice from the windshield, and then driving to work in a car that only got warm and cozy as I was arriving at work.
    I completely agree. Just too much shoveling, scraping and slipping for me. Any peculiar desire to frolic in the snow is quickly subdued by watching the winter road condition reports from the Sierras. Still, I have to admit I have this desire to smack someone with a snowball! }:-)

    Amanda

  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2015-11-16 16:25
    I suspect I have reached an age where slipping and falling on ice might be fatal.

    A good snow depth gage for me would be the number of lost postal workers or paper delivery persons.

    One can always determine the weather is not fit to go outside by asking for a pizza delivery. A refusal indicates one should remain where they are.
  • Duane DegnDuane Degn Posts: 10,588
    edited 2015-11-16 18:52
    Bean wrote: »
    I think the 28044 sensor would work fine. It doesn't do "time of flight" measurement. It simply looks for where the laser dot it appears in the camera.
    Snow should show up very well. I would be willing to bet that it would work quite well.

    I'm not so sure that sensor would work well. It has a hard time outside in the sunlight.

    I have one myself and I think we're due for some snow soon. I just looked out the window and it's snowing now. It's not deep enough to really measure yet but once there's a bit more snow, I'll test a couple sensors on it.

    I personally think an ultrasound sensor would work. I'm not sure why I think this but snow just doesn't seem the same as soft fabric. I think it would reflect the ultrasound. I'll test this out too.

    Speaking of snow, I think I owe erco a figure 8 in the snow.
  • I use the ultra sonic sensor to measure water level in a sump pump well, this works fine for distances below 4 meters. If you get the seconds from ebay they usually work at least two meters.

    They say the Parallax 28044 laser range finder doesn't work so well in direct sun, and I thinkit would work less well in snow.

    The SF02 on the other hand https://www.parallax.com/product/28043 works very well in direct sun to its full 50 meter range, although I haven't tried it on snow in full day light. I should try this next time it snows.

    So a big question is where will you be measuring, will it be in full sun light; and how much snow do you intend to measure, will it be under two meters or will it be over four meters?

    This will tell you which will do the job.

    Then just buy the most expensive one, because that will be the most fun to play with.
  • I would like to find a method that works day and night. I have a Davis Weather Monitor II weather station and use a Propeller to upload sensor data every 5 minutes then transfer these data to a W5200 connected to the internet so the neighbors and I can see the local weather conditions on our various mobile devices. Adding real time snow accumulation is next on the list of things to measure. I am planning on building a 24"x24" platform scale to measure the mass of accumulated snow and would also like to measure its depth.

    Snow has a nominal bulk density somewhere in the 0.07 to 0.12 g/cm³ range (4.4 to 7.5 lbs/ft³). Compare this to acoustical foam which has a nominal density of around 0.03 g/cm³ (1.7 lbs/ft³) and one can see this might be a bit of a challenge. Added to the low density difficulty is the random orientation of the snow surface. A flat surface perpendicular to the direction of the PING beam should reflect enough sound back to the sensor but any significant undulation of the snow surface will send the incident sound scattering off to la la land.

    I have been playing with a PING sensor using various pillows and plush blankets as proxies for freshly fallen fluffy snow. Depending on which blanket and its flatness I can sense its location with decent accuracy and fairly good reproducibility. However, a slightly rumpled surface can often result in no reflection back to the sensor. A very fluffy throw blanket is particularly troublesome for the PING sensor. A couch pillow with noticeable curvature seems to be OK, presumably because it has a relatively dense fabric surface (the fluffy poly fill doesn't get a chance to come into play).

    I haven't come across many documented uses of lasers for snow depth measurements. I imagine the spectral reflectivity of snow is quite intense since it is made up of very fine ice crystals. This might preclude the use of simple laser proximity sensors.

    Perhaps the only way to combat the high spectral reflectivity of light and the low acoustic reflectivity is to throw a high level of energy at the surface in hopes of seeing something reflect back. I have a feeling this is what the $150 ultrasound detectors are doing.

    No snow in my forecast in the next week so I will continue with the blankets.

    I would certain appreciate any feedback from anyone fortunate, or unfortunate, to have snow already!

    PS - I hope a 2 meter sensing range will be sufficient. We had lots of snow last year here in central Massachusetts but the average accumulated depth never got much greater than about 3 feet.
  • kwinnkwinn Posts: 8,697
    Perhaps a deep tube with a floating bottom could block sunlight and allow optical sensing to work.
  • No snow hardly ever here around San Francisco except for dusting on Mt. Diablo and Mt. Tam and Mt. Hamilton! Having been born in Denver and having lived for formative years in Michigan I do miss the snow, sometimes. The snow we get in the mountains around here (less of it recently) is affectionately (or not) call "sierra slush". Quite hard packed, structured layers of ice and wet powder.

    I helped with instrumentation for the USFS Sierra zone observatory, and that used the Judd ultrasonic snow depth sensor. ($$$) The sensor does in fact have trouble during active snowfall, especially when windy. Sloping terrain is to be avoided. Even so, the data set forms a very nice dotted line during periods of relative calm. There is a photo of the hardware and a description at http://criticalzone.org/sierra/infrastructure/sensors-field-instruments-sierra/. The system pictured is based on a BASIC Stamp (long live it!) and a custom sensor interface board and a vDrive USB flash interface, and a Dust Networks mote. Recently a couple of the principals involved involved in the project started their own company and replaced the Stamp logger with their own, Metronome Systems, which integrates the latest Linear Tech/Dust motes along with a fancy Cypress PSoC.

  • Looks like a nicely packaged sensor. I emailed them inquiring about current pricing.

    Does anyone at Parallax know the output power of the 40 kHz sound pulse from the PING sensor? I cannot find it in the online documentation.
  • How about an array of 4 (or more) Ping sensors, on a cross shaped mast that each measure the snow depth. Even if a few are over undulations, you might get at least one good reading. The shortest ping is the highest snow depth, and any off axis reflections will read long or not at all and can be discarded.
  • SeairthSeairth Posts: 2,474
    edited 2015-11-18 07:29
    I wonder if you could use a technique like this:

    http://www.circuitsgallery.com/2012/05/numeric-water-level-indicator-liquid.html

    There are plenty of small ready-made versions of this, but I expect you'd need to make something custom for 2m depths.

    Edit: as an alternative, you might also be able to use a capacitive approach. Place two long strips of copper next to each other, with one connected to ground and the other connected to the Propeller. Then, using the same pulse-then-sample technique I used with the Propeller Badge pads, you should be able to measure how much charge us leaking to ground, which should be proportional to hire deep the snow is. I'm sure there are still some technical challenges to this approach, but the party I described above would be dirt cheap to implement.
  • It seems that as the winter progresses, the snow might just get deeper and deeper and eventually bury a sensor.

    So why not have a pole climbing sensor that recalibrates as it climbs the pole? With an ultrasonic sensor having a range of about 4 feet, one suddenly has the ablity to vastly increase its range.
  • Sapphire wrote: »
    How about an array of 4 (or more) Ping sensors, on a cross shaped mast that each measure the snow depth. Even if a few are over undulations, you might get at least one good reading. The shortest ping is the highest snow depth, and any off axis reflections will read long or not at all and can be discarded.

    That's an interesting idea. One could then trigger all four sensors at the same time and measure the response time for each individual sensor. If the snow is cooperative then you would see four very similar response times and report the average value. If on the other hand the snow is being particularly uncooperative you might "sense" one of the outgoing pings on at least one of the other sensors. If you receive two or more dissimilar responses then you could average them or develop an algorithm to determine the most probable value.

    I got a price from Judd of $700 for their snow depth sensor. The MaxBotix HRXL-MaxSonar-WRS Series sensor is about $150 and comes ready to mount. A set of four PING sensors would be $120 plus the cost of a suitable enclosure. From a pure economical standpoint the MaxBotix sensor seem the better option so far.
  • It seems that as the winter progresses, the snow might just get deeper and deeper and eventually bury a sensor.

    So why not have a pole climbing sensor that recalibrates as it climbs the pole? With an ultrasonic sensor having a range of about 4 feet, one suddenly has the ablity to vastly increase its range.

    If I have to move the sensor up that's my signal to move to another climate!
  • I'm pretty sure the Judd sensor uses the Senscomp (formerly Polaroid) ultrasonic transducers. You might find the Senscomp "ultrasonic ranging" site interesing.
    http://www.senscomp.com/ultrasonic-sensors/complete-ultrasonic-ranging-solutions.php
    Check out the ICs available there too if you are interested to DIY. It is not out of the question. The Propeller by itself would be great for the pulse generation and return echo timing functions.

    Many ultrasonic transducers like the PING have open frame construction that does not lend it to outdoor environs. There are other transducers that are sealed and more environmentally protected. (e.g., Prowave 400EP250) I recall that Parallax had an outdoor version in progress, but I'm not sure what became of that.
  • Here are some that are less expensive...

    http://www.ebay.com/itm/331685483931

  • The TSL1401 line scanner is capable of detecting a brightly colored stick poked in the ground and producing an output proportional to the length. I am assuming it would be capable of measuring snow depth by subtracting current length from calibrated full length. I havent tried this and it would require a light source for night readings but it might be an inexpensive method to try.
  • Ray0665 wrote: »
    The TSL1401 line scanner is capable of detecting a brightly colored stick poked in the ground and producing an output proportional to the length. I am assuming it would be capable of measuring snow depth by subtracting current length from calibrated full length. I havent tried this and it would require a light source for night readings but it might be an inexpensive method to try.

    That is an interesting approach. One might opt for a black stick and look for white snow edge. The photodetector has maximum sensitivity around 800 nm (just beyond visible red) so some type of IR LED might provide the necessary illumination at night.
Sign In or Register to comment.