I separate climatic data (min temp, max temp and rainfall) into two periods (the first period is 30 year data and the second one is 20 year data). After statistical tests, I found that rainfall experienced a change in standard deviation but has unchanged mean. Is this situation possible ? Does mean climate change detected due to a change in standard deviation?
Hi Kenneth,
I am feel honour that you consider my question.
My question, after statistical tests, I found that the change of standard deviation of temperature is statistically significant while the change of temperature average. Can I say this is as an indication of climate change ?
Here, mean temp = average temperature in one period (30 years).
a change in stdev and a change in the mean of temp =average temp.
To detect climate change, people usually separate climate in two periods (the 1st period is usually 30 year period, the 2nd period may be shorter 15-25 years). And then we calculate the average and stdev climate (temp, rainfall etc). If average climate in 1st period different to average climate in the 2nd period, it means that climate change is detected....
Muhamad
I think the variations in STD and similar mean value may caused by the period division of the two periods. If you want to explore the long-term disturbance in climte condition, maybe it is better to divide your two period into more details. I believe the long duration in these two period will overlook their internal varitions, because you found that the differences in STD.
Muhamad,
The answer to your question is Yes. The following paper illustrates the increase in rainfall standard deviation (or variance) in the recent 30 years as compared to the previous 30 years (though they are respected to the same long-term mean).
Intensification of Summer Rainfall Variability in the Southeastern United States
during Recent Decades, J. Hydrometeorology, 11, 1007-1018 (2010).
http://journals.ametsoc.org/doi/abs/10.1175/2010JHM1229.1
Dear Muhamad,
As Hui Wang stated above, of course it is possible. Although your means cannot change much, your extremes and patterns can change. This is already impact of climate change. For a detailed analysis, please check the following paper:
Best regards,
Article Projections of climate change in the Mediterranean Basin by ...
Muhamad, the situation is totally possible, not only the standard deviation, but the distributions as well. This kind of signals in the historical records reflect the climate change trend. One of the interesting signals that I have seen is that the mean stays the same, but the frequency and intensity of heavy rainfall increases. The easiest approach to detect this signal is to look at the number of wet and dry days. So if the mean stays the same and number of wet days decrease, the intensity increases. The standard deviation changes.
Muhamad,
It is possible. Both mean and standard deviation can change through time. In order to make the result more reliable, you may need to divide the whole period into two exact same stages (e.g. 25 yrs: 25 yrs) to recalculate the indicators. If the standard deviation or CV still change from earlier to later stages, that would be interesting, and you could make further examination of the possible causes for the change detected. A couple of reasons may be causing this phenomenon, but you may need to first examine the quality and bias in the data series you are using.
Good lack.
- Badges
- Science topic