Despite meteorological instrumentation often being tens of thousands of dollars, there are actually ways to save a buck or two if you look to get your feet wet in remote sensing. Either you’re like me and don’t have a prestigious grant issued to you by a university or government for the purpose of research, or simply want better quality data without locking yourself into a one size fits all “professional grade station.”
There are a few simple ways to lower the cost of quality weather data.
Firstly, know what you want out of your automated weather station. This will largely dictate what you purchase and configure to your liking. As with anything, ensure you know what you want before you buy something, and do the research to be certain what you are doing is what will be best suited for your applications long term. This could include but is not limited to sensor type and accuracy. Understand and distinguish between wants and needs. In this sense, determine what parameters you absolutely cannot live without and those you can. Try to be conservative with this. I know myself personally would love a mast with five different anemometers atop, and three sets of temperature and humidity sensors at three heights. Buuut, that’s not the most practical from a financial standpoint and otherwise. Seriously ask yourself if it is needed. Do I need five anemometers? No. Having a single highly reliable one will suffice. Of course with that being said, everything is relative. Your budget will differ as will your applications. Know what you want from the beginning, do the research, don’t go overboard.
Secondly, determine if what you are looking for is already available on the market. If convenience and wireless connectivity is more of what you’re going for, I’d recommend going with a Davis Instruments or RainWise station. Both are bulletproof reputable brands whos products have been known to last well over a decade. If you are looking for a wired alternative, I’d recommend Peet Bros. Also a very reputable brand that’s made in America. Davis Instruments also offers NIST calibration of their instruments if that’s something you’re into. While the displays of these weather stations are somewhat outdated by todays standards, connectivity has only improved. I would absolutely recommend these weather stations to someone that has a fair budget but does not want to delve into the complexities of remote sensing.
For those that are religious about ensuring the data collected is of utmost accuracy, then consider yourself in luck. I will say, there are some sacrifices that are to be made if you go this route. That’s not to say your data will be of any lesser quality, but some aspects of connectivity and or user friendliness may decrease.
Going down the route of research grade meteorological instrumentation, you open the door to true complete customization. Your data logger, wind sensors, temperature, humidity, and barometric pressure sensors are all independent of one another. This applies to all sensing devices. The beauty of this is that one can choose what sensor they want with what accuracy rating.
The only caveat to all of this is that when deciding what sensors to use, only certain devices I would recommend be purchased second hand. This is because of the inherent nature of accuracies and the sensor’s need to be calibrated for utmost performance. This ultimately comes down to ease of use when it comes time to calibrate. Unless you have the facilities to do so, I would not recommend purchasing temperature/humidity probes second hand, especially discontinued models as support quickly or has diminished for them. The obvious result of this is you’ve spent money on a device that will slowly drift with no way to correct it’s tendencies.
From what I have seen, barometric pressure sensors have a longer lifespan that other sensors. Companies seem to support these past their point of discontinuation. R. M. Young is among the many that do this.
Data loggers are a mixed bag. eBay has a flooded market full of retired data loggers listed from old networks. Most commonly being CR23X’s and CR10/CR10X’s. These data loggers are manufactured by Campbell Scientific and have since been retired from serviceability. They still offer calibration for the devices but will not repair them. I would personally not recommend the CR10X because of it’s small memory size. Unless it is the 2M point extended memory type, I wouldn’t recommend for data intensive applications. Don’t even bother with the CR10, it’s the predecessor to the CR10X and is too old with too little memory to offer. Unless you are logging few hourly parameters, I would suggest something larger to either option. The CR23X is a pretty solid logger. At face value, it has far more to offer than the CR10X. More memory, 1,000,000 data points, and more input channels. This logger also has the ability to house sealed lead acid batteries to negate the need for a stand alone battery and charger. These older loggers use a programming language called Edlog. Campbell Scientific offers PC400 which is a free support software for all of their loggers. This software is free and has CRBasic, Edlog, and Short Cut to offer. I will warn anyone who does not have prior experience with Edlog that it is not intuitive. If what you need configured is simple, consider only using Short Cut and modifying the programs through Edlog. LoggerNet is another option, however listed at almost $800, it is quite the commitment. I’d recommend doing your research on either software provided by CSI to determine what will best suite your application.
At the time of writing this article there are several CR23X loggers listed for around $300 on average. The CR10X also usually goes for this much, though keeping tabs on listings can usually yield one for sub $200. For what you get I think the price is quite inflated. Unless it’s the 2M point variant I would not bother. The CR23X on the other hand, I wouldn’t be opposed to this price.
Something VERY important to note with the older data loggers is the CS I/O port on them. This looks like an RS-232 DB9 connection but is anything but. The CS I/O port is a proprietary system used by Campbell Scientific that has a different logic level than RS-232. Because of this you are required to use an SC32B optically isolated RS-232 converter. Unlike the CR10X, the CR23X has an additional true RS-232 port not requiring the use of the SC32B. While this is nice, some devices require the RS-232 port or CS I/O port to be connected directly. Take that information with caution and consideration while deciding on a data logger.
Oh yea, the CR21/CR21X. Don’t. Don’t even waste your time looking at the listing. Over priced ancient tech. Most listings are overpriced and what you get is glorified retro hardware, going on 40 years old now. Unsupported and ancient I/O. Unless you have an extremely light data demand and simply setting and forgetting it, I would not recommend.
CR10/CR10X, CR21/CR21X, CR23X, and CR500/CR510 are all older data loggers retired by Campbell Scientific. The first five are most commonly available. I have yet to see a CR500/CR510 listing, those are good for small data use cases. I would not recommend purchasing anything newer than these loggers from eBay as you don’t get the benefits like warranties and service from CSI themselves.
While buying new is sometimes a pain, especially financially with data loggers, doing so provides greater protection for the device directly from the manufacturer. Not only this but you have all of the new capabilities that comes with a new logger. If ones budget allows this, I would recommend from the start. Most have the option for built in connectivity including WiFi and communication over cellular networks.
Like stated prior, if you’re like me and on a pretty strict budget, acquiring a logger from eBay is more than enough. A significant portion of money can be saved going this route. Purchasing any one network interfaces from Campbell Scientific lets you communicate via other means than directly over cabled line. These run around $500 or so but completely renew your logger’s capabilities. Going this route nearly halves costs for data acquisition and frees up budgetary spending for actual sensors. Though this is of course anecdotal and estimated at best and basing off my experiences. (Your mileage may vary.)
Now with data logger understanding out of the way, sensors can be addressed. For the sake of simplicity, I will work my way down as if I am referencing a standardized meteorological mesonet tower.
Wind sensors. Depending on what type, mechanical or ultrasonic, can be bought second hand and calibrated. The industry standard for wind measurement is the R. M. Young Wind Monitor. There are several variations of this device and comes in different output types. Most commonly used and deployed throughout the world is the 05103 Wind Monitor. This configuration is well known and will connect to your data logger with no issue. Something very important to note with these wind sensors is their output type. The “default” configuration is the standard output, which is a pulse for the wind speed and single ended voltage for wind direction. Loggers have no issues here. The second type of output is 4-20mA. This type is common with industrial devices and requires a 100 ohm 0.01% tolerance resistor bridging inputs to allow for differential voltage. Very easy to configure and works with most if not all loggers. The last output type is 0-1V and 0-5V voltage out. Each voltage range is proportional to the full envelope of range the sensor has, each corresponding to 0-1V and 0-5V. The CR10X has a maximum voltage input of 2500mV or 2.5V.
Because of this, changing the wind monitor output to 0-1V is required. The wind speed multiplier is set to 0.1. The wind direction multiplier is set to 0.355. Or alternatively one can use the full 0-5V output but doing so requires a voltage divider from CSI to halve the 5V to 2.5V. Either or, the sensor will work with the data logger. The CR23X has a maximum voltage input of 5V so no modifications are required. This knowledge can be applied for all Wind Monitors.
Calibration can be done either by oneself or R. M. Young. They still calibrate wind monitors from the late 1980’s because of how outstanding their design is. Completely reputable and long lasting.
Ultrasonic anemometers are a bit different. Unless the sensor is being calibrated by the manufacturer, I would steer away from these unless it is new in the box and has compatible outputs with older data loggers. (Most of the time they will.) If it does not look perfect in a listing, I would not even consider it. Given the transducers have to be perfect to function, there is more risk involved with these unlike mechanical style anemometers.
Temperature/Humidity sensors are similar to ultrasonic anemometers. Unless they are still being manufactured and supported, there isn’t really a point in using them. If you have the means to calibrate them then by all means go for it. The only problem I have found with second hand sensors is because they are unable to be serviced, getting replacement parts for the humidity sensor component is next to possible. That is really the only big problem point there. Because the little chip within them can’t be replaced, the eventual drift will render the sensor useless for accurate observations.
Barometric pressure sensors (assuming they are still supported by the manufacturer) are also something to be bought second hand. Unless drift is really bad, recalibration is possible. Even if a manufacturer no longer supports the sensor, I do believe NIST calibrates all for a fee. I am not one hundred percent sure on this but am more than certain this is a capability of theirs, regardless of sensor.
Rain gauges are similar to anemometers with regard to lifespan. Their calibration is relatively simple and parts for replacement are easy to come by.
The sensors and data loggers highlighted are the big ones in my opinion. As such with other sensors, gauging their serviceability and age will most likely give you an idea as to how one should approach the purchase. Hopefully this information here provides some insight for those looking for information on alternatives as a means to save money. Sure little sacrifices may be made now but in the long term the money saved for functional equipment cannot be ignored.