Hello, I want to calculate the percentage change in rainfall by taking the future rainfall values and the baseline period values. My baseline period is 1961-1990 and future values are 2041-2070 and 2071-2099. I want to know this baseline period values should be taken from the observed rainfall data or model simulated one for the same period?
Thank you in advance.
Hi Asfandiyar,
Climate scenarios developed for impact applications typically need some estimate of climate change to be combined with the observed data. The more observed data (both spatially and temporally) the better one can define the baseline. Therefore, this is why there is a high demand on acquiring more sophisticated observational data in recent years.
Also, note that choosing a year of data for defining the baseline is not appropriate because that year might be too cold, dry, or wet. Typically 30 years of data is preferred as it can capture all inter and intra-annual fluctuations (between years and seasonal). When the observational data are sparse and limited, defining the baselines may become less accurate and less reliable.
Hope this helps you
Best regard,
Ebrahim
Hi Asfandiyar,
There could be differences between the 1961-1990 climate in the observations and that simulated by the model/s. Therefore, it is not safe to get a ratio between, say, the 2041-2070 mean precipitation in the model/s and that in the 1961-1990 in the observations, as the observations and the simulations might not be representing the same climate.
My advice would be to compute the difference between the future period and the 1961-1990 period in the model/s (if it is more than one, do it individually and you may then eventually average the differences) and then apply that ratio or coefficient to the observed 1961-1990 values. Still, this approach has a limitation, since it considers that the deviation in the present climate between obs and model will remain the same in the future, which might not be true, but is better than mixing obs and model across the different periods.
Hope it helps.
Cheers,
Ramiro
Last time I checked ( a few years back) there were about 10 climate change models based on a variety of assumptions. One can make various calculations with data and models, but to cite some of the comments perhaps taken out of context, such as it is not safe, it is not appropriate, less accurate and less reliable, and might not be true pretty much sums up my impression of predicting changes or even averages 20-80 years into the future, based on period data 30-60 years in the past. One can pretty much make any assumptions, calculations, and projections we might want to and there would be no current ramifications probably on us personally if we were totally wrong, 20, 50 or 90 years into the future. But some people rely on scientists and professionals, and if there are some who might depend on the projections, not knowing the assumptions and weaknesses, they might be hurt by relying on such futuristic extrapolations. If nothing else, qualify the data, methods, results and uncertainties to help avoid misuse or unwarranted expectations.
The base line data should be observed or measured rainfall. Observed rainfall is normally used for calibration. So you can find the mean for the base year and compare this mean with what you get in future.
In my view the model can simulate various scenarios based on different variables, but baseline data is sacrosanct and should not be meddled with.
- Badges
- Science topic