GaugeMap: Thousands of river gauges, each with a Twitter account • Hacker News Board

Re: GaugeMap: Thousands of river gauges, each with a Twitter account

#31

Post by blakecaldwell » Thu, Dec 31, 2015, 5:21 AM UTC

Animats wrote: post #19

olva22 wrote: post #4

This is amazing! I've wanted to do some "citizen science" project in waterways around me. For instance, building cheap sensors that detect temperature/water level/salinity/particulates/etc that are so cheap as to be basically disposable. The sensor itself could probably be made for less than $10 these days using various arduino modules, but I'm stuck on how to have the sensors communicate back to a "base station" to…

Search for "M2M" (machine to machine) communications. There's a lot of this going on over old 2-way paging networks. It's how the non-vaporware "Internet of things" works. There are thousands of air conditioning compressors, refrigeration units, elevators, pumping stations, and other boring but important pieces of infrastructure sending their short messages every few minutes. Often they're plain text, something like…

Satellite is also an option for remote locations:

https://www.sparkfun.com/products/13745

$250 for this module, and $12/month for the satellite usage.

For the sensor itself, that's a simple micro controller for under $10, and a $5 sensor, a battery pack, solar panel, and a most likely, a few resistors. Costs add up quickly if you don't want to run out and replace the batteries, and if you want it to phone home.

Re: GaugeMap: Thousands of river gauges, each with a Twitter account

#32

Post by thisnews » Thu, Dec 31, 2015, 5:46 AM UTC

syedkarim wrote: post #26

thisnews wrote: post #22

Earlier quoted context omitted.

I am a hydrographer in the NT working out in very remote areas- Phone reception is rarely possible for such monitoring sites. Currently we use satellite Comms. to relay water data home, but this is very expensive and prone to failure. I'm wondering if you, or anyone else, has some ideas on long range communication strategies for large areas (NT is 1.4 million km2). The area we are required to cover generally means si…

Which operator are you using? We've found L-band to be very reliable and does not suffer from any of the attenuation issues found in Ku. And heavy spreading on the return path provides for an even more resilient link. Disclosure: We'll be offering an L-band satellite gateway for IoT this summer.

We currently use Iridium. On Ku band I beleive. I'll look further into this and see what our options are for L-band.

Outernet looks like a great project. How small of a satellite can be used to achieve the connection? Any more info/documentation on the L-band gateway?

Re: GaugeMap: Thousands of river gauges, each with a Twitter account

#33

Post by RiverKey2640 » Thu, Dec 31, 2015, 8:53 PM UTC

olva22 wrote: post #4

This is amazing! I've wanted to do some "citizen science" project in waterways around me. For instance, building cheap sensors that detect temperature/water level/salinity/particulates/etc that are so cheap as to be basically disposable. The sensor itself could probably be made for less than $10 these days using various arduino modules, but I'm stuck on how to have the sensors communicate back to a "base station" to…

I've been supporting hydrological data collection for over two decades. When I first started in the early 90's, there was no clear defined trail on 'how to do it'. I learned over time what works and the right way to do it. I'd like to make a few comments.

You asked about organizations that engage in 'citizen science', I would start with the NOAA Co-ops program https://www.ncdc.noaa.gov/data-access/land-based-station-dat... . This would give you the basics. Another route taken is through a college based program. http://iowafloodcenter.org/projects/stream-stage-sensor/rive... As you can see, Iowa has a program that is on the lines of what you are looking at.

You mention about cheap 'disposable' sensors. Truth be told, no reliable sensor is truly cheap. In sensing, we look at repeatability, accuracy and resolution. In the design of a sensor we also look at interfacing and field calibrations.

Field data collection systems are comprised of several parts. Power, Sensing, Data Logging and Telemetry. It's interesting, different vendors of this equipment take different approaches. Everything revolves around the abilities of the data logger. Let's look at several:

Hobo: http://www.onsetcomp.com/products This is a pretty good company. The sensors have good accuracy. The products are reasonably priced. However, you are locked into their products and limitations. I've seen Hobo used by students in course work.

What does the USGS, USACE, NOAA and BUREC use for data logging? They use capable data loggers that can accept SDI-12 and ModBus based sensors along with Analog based sensing. CS: https://www.campbellsci.com/dataloggers and Sutron: http://www.sutron.com/product/xlite-9210b/

As brought out in this thread, Telemetry is the weak link as the trade offs are range verses bandwidth verses data charges, etc.

The notion of Real-Time data is also a factor. Is Real-Time 'right now', every 15 minutes or hourly? All play a factor in telemetry.

@RiverKey2640