Can I do something scientific with daily temperature data?

[GhostLoveScore]

I've been logging daily temperature and humidity data for two years now. In the end of each year I will put a graph like this

Those are daily minimum and maximum temperatures. But I was wondering, if all this has any scientific value at all. I've been using WS2300 weather station for these measurement and it seems it's accuracy is only 1 degree C.
So,I've been researching other temperature/humidity sensors and came upon this one http://ww1.microchip.com/downloads/en/DeviceDoc/25095A.pdf MCP9808. It has typical accuracy of 0.25 degree C and I bought two. I was thinking if I was using two sensors and calculating average temperature from two together that I would get 0.12degree C accuracy. I don't know if that's true.

Back to my question, does this what I'm doing has any value whatsoever? What can I do to make it better?
And I have logged temperature for each hour every day, so I have that too, I don't know what to use it for.

 

[CWatters]

The UK Met office accepts weather data from anywhere in the world...

http://wow.metoffice.gov.uk

 

[.Scott]

You could compare it to the local temperature records. Your very local geography (within a mile of your instrument) may have an affect on temperatures - perhaps making them slightly higher or lower than other stations near you. Or there may be temperature affects related to wind speed and direction.

In many areas, you can contact your local TV meteorologist and arrange to provide them with frequent temperature reports.

 

[fresh_42]

You can use these data as an example to study several techniques in probability theory and statistics or differential geometry. I cannot really imagine they would be of any scientific use other than an example. For climate considerations it is far too local and for weather considerations not long enough. You have very local information in space as well as in time, which therefore only allows you local investigations like the underlying distribution, and related quantities or of the vector fields you created.

 

[OmCheeto]

Back to my question, does this what I'm doing has any value whatsoever?

As a homeowner, I would use such data to optimize a system that would minimize the thermal/financial depletion of my bank account.

 

[GhostLoveScore]

For climate considerations it is far too local and for weather considerations not long enough.

How long would you say I need to monitor the weather to have something I can analyze?

 

[fresh_42]

How long would you say I need to monitor the weather to have something I can analyze?

The problem is not that you cannot analyse it. The problem is, that you have to compare it to something. Meteorologists usually have data over decades or even centuries, depending on location, and the development over time is interesting rather than the observation of a few years. By the current instability of our climate I assume that you can already find significant changes in variance, extrema or similar quantities within - say 10 years. This will not allow you to extrapolate global data, but can show local differences to long term means. E.g. we had a record breaking summer in 2003 in central Europe, and I think again two more recently. Thus such events can be observed even locally, but to know, that their frequency has actually changed, you need to compare it with older data over decades and centuries. Maybe those long term data are nowadays easily available at where you live, in former times, people had to study chronics which are kept in churches.

I still think it's a great example to study the various stochastic concepts, and in a second step compare them with other data of your area. You can really, really, learn a lot about statistics and feasible forecasts. However, I have to warn you: you will never again be able to listen to a politician who uses data as an argument!

I also found these two comments very interesting:

The UK Met office accepts weather data from anywhere in the world ...

... to provide them with frequent temperature reports

I know that we have a similar system here, and that there is one in Russia, where people contribute their daily data to the often governmental office of meteorology, especially if they live in far places like mountain tops or arctic regions.

 

[GhostLoveScore]

The UK Met office accepts weather data from anywhere in the world...

http://wow.metoffice.gov.uk

Thanks, I'll try it.

You could compare it to the local temperature records. Your very local geography (within a mile of your instrument) may have an affect on temperatures - perhaps making them slightly higher or lower than other stations near you. Or there may be temperature affects related to wind speed and direction.

In many areas, you can contact your local TV meteorologist and arrange to provide them with frequent temperature reports.

Unfortunately I can't do that, we don't have any local TV stations except large ones that are located in capital, some 150km away.

 

[GhostLoveScore]

As a homeowner, I would use such data to optimize a system that would minimize the thermal/financial depletion of my bank account.

Well, the system here is that you put wood in the fireplace, so I don't really know how I would optimize it :)

 

[fresh_42]

Well, the system here is that you put wood in the fireplace, so I don't really know how I would optimize it :)

You can calculate whether solar cells are a good or a bad investment: Your data can possibly give you the amount of sunshine, or you can estimate it from the data (compare humidity and temperature on sunny days in winter and summer and read your data to estimate the total amount), then consider degree of efficiency of solar cells, costs and earnings.