Purchase of a weather station

The years 2020-2023/2024

The BMCB network currently in operation consists of about a hundred Barani Design monitoring stations. Its main asset is the Meteoshield, which acts as a "miniature thermometer shelter" and, despite its small size, provides very good measurement results, comparable to those of its big brother, the Stevenson screen. Moreover, the station operates completely autonomously, which means it can be installed in open areas, away from disturbing influences that can make measurements less reliable.

The network has expanded significantly in recent years, from about 10 monitoring stations in late 2020 to nearly 100 in early 2024. This also means that the oldest stations are now three years old. Unfortunately, all the weather stations installed in 2020 have also broken down over the years.

It must be said that the number of problems has increased significantly in recent years, which is why we decided to stop offering these BARANI weather stations. You can read more in the section “Measuring equipment”.
 

Our solution: the development of a new weather station

Because of the persistent problems and no prospect of a good solution we have been working for the past year on the development of our own, new weather station "WEATHERWISE" (abbreviated to "WW")

For us, a number of points are important that we keep well in mind during its development:

1. Reliability of the equipment: we strive to have no defects and if any, to fix them quickly. The real cause of the problem must be addressed to avoid it in the future.
2. Accuracy of the measurements: as far as temperature is concerned, it must certainly meet the WMO standard, for other parameters we try to meet it. This usually goes hand in hand with the type of sensors used.
3. Modularity and flexibility: the number of sensors that can be connected to 1 module is limited, but one weather station can contain multiple modules - for example, one module for temperature, humidity, and precipitation, and a second one for wind speed and direction. We also provide multiple interfaces (optional) for additional sensors and also support existing (semi) professional sensors from known manufacturers.
4. Cost: of course we want to keep it as low as possible. However, we do not want to achieve the low cost price by using inferior materials, taking risks, or compromising on the points mentioned above, which are more important to us.

Things that are less or even not important to us: small size, aesthetics...

The first version of the developed weather station is currently being thoroughly tested, and needless to say, outside in the elements. The test sites are located in places where the weather conditions in our country are the harshest for electronics: low temperatures, little sun, greater humidity for extended periods of time, high precipitation, snow... The Ardennes are consequently the ideal environment to thoroughly test the new stations. The locations we chose are Elsenborn/Bütgenbach (Grünen Kloster) known for its sometimes very low minimum temperatures, Losheimergraben/Büllingen near the German border because of its high altitude and in a forested area, ideal for little sunshine, as well as the locations Braunlauf/Burg-Reuland and Recht/Sankt-Vith.

Winter season is almost over. On January 20, the station in Grünen Kloster/Elsenborn recorded the lowest temperature in the past 4 years: -23.0°C and just above the snow layer a respectable -26.2°C. It was the WW test station that allowed us to measure this.

During this test period, two shortcomings were discovered that have been eliminated in the new (modified) version, which will be rolled out soon.

The new weather stations also use the Sigfox network like the Barani Meteohelix and also transmit data every 10 minutes. Possibly other network types such as LoRa will be added in the future.

Where a Barani Meteohelix had no ability to measure temperatures near or in the ground we have also added this functionality. Because we control all the steps of the new design, we can go in any direction when demand arises.

Specifications of the new weather station

Below you can find a first overview of the new weather station with also a comparison to the Barani-Meteohelix/MeteoWind/MeteoRain

Specifications of the WeatherWise weather station and comparison with the Barani Meteohelix
Barani "Meteohelix - Meteowind - MeteoRain"	BMCB "WeatherWise"
Compact weather station where all sensors (except rain gauge) and electronics are located internally in the Meteoshield.	Somewhat larger weather station consisting of a plastic housing (polycarbonate, IP68) containing the electronics and lithium battery. In addition, a solar panel is present and for most people also a (Barani) MeteoShield as a Stevenson screen (in other words, what is good we do continue using...).
Measurement of air temperature, up to 0.1° accurate for range +20..+60°, beyond that 0.3° according to sensor manufacturer (this contrary to what Barani states in its specification); average, minimum and maximum temperature are also available	Measurement of 2, 3 or 4 temperature channels each within 0.15° accuracy (without sensor calibration), well better than 0.1° after calibration. For the version with 2 or 3 temperature channels, data of average, minimum and maximum are available, for 4 channels only average and minimum of all channels and only maximum of 1st channel.
Measurement of relative humidity, accurate within 2% to 90% RH, above that possibly greater inaccuracy.	Development of the separate measuring probe is in progress; the probe itself is in stainless steel. The sensor used for temperature and relative humidity is the same as the Meteohelix. Consequently, the specifications are the same as MeteoHelix.
Measurement of atmospheric pressure. Sensor accurate +/-1.5 hPa at 25°, beyond that increasing to more than +/-3 hPa for the temperature range -20/+50°C	A separate air pressure (barometer) station will be available, accuracy better than 0.30 hPa, but we are aiming for 0.15 hPa (but is yet to be confirmed).
Measurement of solar radiation via their solar panel; according to Barani specification class C but this is not achieved at all (after measurement with class A pyranometer) and over time this gets even worse.	An additional module is being worked on that will also allow measurement of solar radiation with a (real) pyranometer, supporting many types. A good measurement needs a good sensor and we get these from specialists in the field.
Measurement of dew point: is not measured but calculated from air temperature and relative humidity. This is good for RH below 90-95%, above it the deviation increases, mainly due to the larger measurement error of the humidity sensor above 90-95%.	Same as Barani MeteoHelix.
Precipitation: connection for a switch (reed switch) of the rain gauge (tipping bucket principle). However, experiences with the Barani rain gauge are not good; many devices show large deviation.	Precipitation: connection for a switch (reed switch) of the rain gauge (tipping bucket principle), so same principle to the Barani MeteoHelix system. We are looking at which rain gauges perform well and will make a choice from them. We will not develop a rain gauge ourselves.
Wind: Barani MeteoWind station (note: does not come with Meteohelix and is a separate item to be purchased) with average wind speed, wind peak at 3 sec, wind direction, wind gust direction.	Wind: same regarding housing, electronics and solar panel; we are now using the Barani MeteoWind Compact sensor (so not the IoT version). We process average wind speed at 10min, wind peak for 3 sec, wind peak for 1 sec, wind peak instantaneous, average wind direction 10 min, standard deviation wind direction and wind speed. Our weather station also supports wind sensors from other "major" manufacturers as well if Barani wind sensor is no longer available at an affordable price.
Network: Sigfox, integrated transmitter with small internal antenna which does not have the best efficiency; we also noticed that the direction in which the MeteoHelix is set up determines the signal strength.	Network: Sigfox, integrated transmitter with an external antenna. An external antenna (usually) gives better results and depending on location (network coverage) a small or larger antenna can be used. The antenna can also be installed higher in locations where Sigfox coverage is slightly less.
Solar panel integrated on top of the MeteoHelix; after 3 years of experience we find that the solar panel used is of low quality.	Separate solar panel. Disadvantage: it is bigger. Advantage: in less sunny locations you can just use a bigger panel. Solar panel is weatherproof with a projected lifespan of at least 10 years.
Very compact device.	Less compact than Barani MeteoHelix due to separate parts and larger solar panel, slightly more wiring (depending on how many sensors are connected).
Not possible to use other temperature sensors.	Additional module (optional) is being developed for the very stable PT100/PT1000 temperature sensors (platinum resistance sensors).
No memory. If the transmission/reception of data goes wrong, data is lost.	Each module has memory to store data of a full year, including some additional data. For example, temperature is available for every 2 minutes whereas via Sigfox this is only transmitted every 10 minutes. This, however, does require a cable connection.
Recalibration only possible afterwards on server.	Each module has the capability to calibrate the sensors (where applicable); this calibration is stored in the module and the transmitted data are thus immediately correct. No mess with corrections afterwards.
Not possible to attach sensors using Modbus RTU communication.	Additional (optional) module is being developed for connection of sensors equipped with Modbus RTU communication (e.g. certain types of pyranometers and weighing rain gauges use this protocol).

In the near future you will hear more information here about results we obtain, as well as problems encountered and how we intend to solve them. We like transparency and strive for good quality and reliability of the weather station in combination with good (accurate) observations.

Furthermore, we want to keep you posted on when we will start rolling it out. We are aiming sometime in the spring of 2024.

More information?

If you have any questions, let us know via our email: [email protected]

Last update: March 19, 2024