Is there anyone interested in hobby electronics to make hobby weather stations? The idea is as follows - with a small amount of electronic components, to build a station in a home environment, which is a sensor generally connected to a microcomputer or desktop with a suitable component. This data is sent to other computers, similar to nodes and/or to servers. From there, the information is used by end users via applications for phones/tablets and computers.

The motivation for this is that nowadays we look at our phones/smart watches or bracelets every day to check what the weather is like outside before we go out. The applications take average data near us, but it is not clear how accurate they are, as they can differ by up to 2 degrees in temperature. For humidity, the difference can even be much greater. If we have better accuracy for the current weather, we could better assess what to do in the next few hours. The same applies if we have to travel in the next few hours. For example, to the villa whether the temperatures have dropped and whether it is urgent to put away the flower pots. By placing a small weather station, it will inform us about the circumstances in the vicinity, and this data will also be used by your neighbors who do not have such good knowledge in the field of electronics, using only the final application.

A few more technical details that I have in my vision of how to make it happen. This is not a final option and I would be happy to discuss.

• the protocol is free and publicly available
• the applications are open source and publicly available
• decentralized data collection
• a small budget for manufacturing with reliable sensors. I have experience with DHT22 whose range is from -20 to +80oC
• selection by data region
• ability to view old data
• ability to be instantly informed about remote data
• a helpful tool if the academic community is interested in the data. For example, in the fight against global warming.

Decentralized data collection can be a mix of releasing packet information through nodes and in combination with servers. In such an option, it would be useful to release packets similar to a blockchain, where there are three significant components in the chain:

• a signed packet (hash) with the sender's private key, which includes the following two components
• a hash with data included in it - date, time, location, temperature, humidity
• a hash key resembling a Merkel tree. The purpose of the hash is to protect the data from spoofing or 51% attacks. Nodes, in synchronization with servers, determine their hash, which changes every minute, and the geolocation of the nodes should not be greater than 50 kilometers in densely populated areas and vice versa - 200+ kilometers in deserts.

When a packet arrives at a node/server/ it stays in memory, from which the location data is extracted, checked for compliance with the Merkel hash and if approved by the node it is optimized for providing the data to applications on request. Most often, it is expected that the memory contains data from the last 10-15 minutes, in order to compare it with other nodes and determine an average value and possible temporary discrepancies. For example, the sensor is temporarily exposed to elevated temperatures, such as direct sunlight in addition to a dark-colored base that heats up excessively. After the sun shifts, the values ​​would be more reliable. With three+ sensors in one area, such as a neighborhood in a densely populated city, they could ignore the one with a high peak in values.

I would even set an alarm on my phone if the humidity rises and the temperatures drop sharply. In my city it is a prerequisite for frost and I would take appropriate measures and I would like to wake up and close the car, pick up the flowers, get up earlier to shovel the snow.

Would you help to build such a hobby network? What do you think about the idea?