I don't know much about them either. I do not think they need to face a certain way. 

I am pretty sure they can be bougt in bulk. That is what the outdoor clothes companies do. 

Try contacting Dale Atkins from RECCO.

Hi Lars, Nida,

Riffing on the Recco idea, if you have access to the detector, it's possible to make a Recco-compatible reflector for very cheap, since for this purpose it won't need to be sewn into clothing or put through a washing machine.  There is no magic behind the reflector technology, I think it can be made in small quantities for USD1-2/each.  It is basically RFID technology, and you know how cheap they're trying to be.  Here is a paper talking about just that.  Especially take a look at their reference [4], which details an attempt at making Recco-compatible reflectors, but using older components (the paper was published in 1997).  There are much better components now.

I bet Recco detectors are expensive.  But there's a way around that!  434MHz transmitters and 868MHz receivers are easy to get; together, these constitute the detector.  Another possibility is 13.56MHz/27.2MHz.  Not sure about the detection range though.

Thanks,

-harold

Hello Erick

A few comments that you may find useful.
Before I start, I am not a complete Arduino beginner but I am certainly no expert either.
I quote British suppliers' websites below but you should find some local to you that offer the same components. If in doubt you can always try Mouser or Digikey (though mind the delivery charges for low value orders).

ARDUINO
1. SIZE - If you want to keep things small and do without a lot of the connectors etc. that come with the "standard" Arduino UNO then have a look at the ARDUINO PRO MINI. A postage stamp size unit with plenty of digital and analogue input/output as well as onboard voltage regulator to give you some flexibility on power supply.
Having said that, you already found out that by default Arduinos are not that energy efficient. But there is help at hand if you don' t mind getting a little more immersed in the electronics. See here for a very useful and detailed tutorial that worked a treat for one of my projects:
http://www.home-automation-community.com/arduino-low-power-how-to-run-atmega328p-for-a-year-on-coin-cell-battery/

2. OPERATING VOLTAGE AND CPU SPEED - There are Arduino Pro Mini versions operating at 3.3 or 5V, 8 or 16 MHz. Unless you need lots of processing power (unlikely from your description), then the 8 MHz version will do perfectly well. Pick the voltage version depending on the peripherals you are going to use. There are usually, but not always, 3.3 and 5V versions. SD card readers, for example, operate at 3.3V and those suited for a 5V Arduino have onboard level shifters to convert voltage levels.
http://www.hobbytronics.co.uk/index.php?route=product/search&keyword=arduino%20pro%20mini

3. SAVING ENERGY - I guess you don't need continuous measurements but just, say, once an hour. In so far, you could use slightly more energy-intensive measurement devices as they (and the Arduino) will only ever run for a few seconds at a time. On that note, you should be able to power all or at least some of the peripherals off output pins of the Arduino. They only supply a few 10mA but this is enough in most cases. Use two outpins per device, one set to LOW (for 0V) and one set to HIGH (for 3.3 or 5 V). Larger devices would have to be switched on and off through a transistor or even relay that is controlled by the Arduino.

REAL TIME CLOCK
Beware of some of the cheaper RTCs. I found to my detriment that, e.g. DS1302, does not keep time well at all to the point of being entirely useless. Better spend a few bob more and get something like the DS3234, which worked fine for me.
http://www.hobbytronics.co.uk/rtc-ds3234?keyword=rtc

MEASURING WATER FLOWS
If at all possible, measure contactless. As soon as you get your sensor anywhere near the actual stream, you are exposing it to all manner of potential trouble including debris getting washed onto it, clogging any mechanical sensor and possible upsetting most other "direct contact" sensors. Having said that, there don't seem to be all that many off-the-shelf options so here are some ideas off the top of my head.
Digikey sells the same sensors previously suggested from Adafruit but they also have a range of other "immersible" sensors that are lot cheaper:
https://www.digikey.com/products/en/sensors-transducers/optical-sensors-photointerrupters-slot-type-transistor-output/548?k=optical%20liquid%20sensor
A waterproof capacitive sensors may be an option but this would have to be helddown or burried on the stream bed:
https://www.ebay.com/itm/DC-5-24V-Contactless-Liquid-Level-Sensor-Switch-Non-Contact-Waterproof-Detector-/122834485435
An entirely different approach but definitely non-contact: use a small camera module and machine vision to detect the change of state between dry and wet. To do the processing of this on an Arduino would likely be a challenge, then again, what you are looking for is quite a simple change in visual appearance that may be possible to detect without too much processing.

POWER SUPPLY
If you are operating in the desert then I expect a small solar panel with battery should provide you all the power you need for such a small unit. I can't give you much more details on something custom-made but if you want to keep things simple then use an off-the-shelf device with a USB-style connector (get the 5V Arduino in this case). Better yet, use something that has something more robust than a USB port. If all else fails, use something like Sugru to beef-up the USB socket and plug against accidental damage and disconnection.
https://sugru.com/

Good luck and do have fun !!
J