Research into the interlinks between biodiversity patterns, processes, threats, conservation and human welfare



CLARIFICATIONS



1. After our paper was published in Nature, it was pointed out to us that some previous studies, although based on different metrics, had reached similar conclusions about the prompt emergence of unprecedented climates see here. All these papers combined help to consolidate the main finding that climate change is happening and will reach levels of significant concern very soon. I, Camilo Mora, also publicly acknowledge that this paper was possible thanks to the dedicated work of climate modelers around the world. They were acknowledged in our paper, but they deserve a huge amount of credit for unselfishly providing public access to their data, making all these secondary analyses possible. This paper is theirs as much as it is ours.

2. The paper provides two measures of variability that we explain in detail here: 1) geographic variability (i.e. variations in the timing from place to place) and 2) the error in our index (i.e. precision or variations among models in the timing for any given location).

In our paper, we report +/-14 years (standard deviation) for the global average of 2047. This represents the geographical variability from place to place, which is a key point in our paper (that is, unprecedented climates are not happening at the same time everywhere). Just to give you an example, the timing of climate departure for Santo Domingo, Dominican Republic is 2026 and for Kyoto, Japan it will be 2040. What was remarkable about the analysis was the consistency of the results among models. There was never more than 5 years standard error on any given place. So in other words, there is little uncertainty in the values for those locations; the precision among 39 models was no more than 5 years of standard error (see result at Extended Data Figure 2a in the paper). Adding the standard error to the projected multi-model average timing for Santo Domingo will lead to unprecedented climates emerging between 2021 and 2031, while for Kyoto the timing will vary between 2035 and 2045.

Below we provide an illustration of how to interpret the standard deviation.
Basic definition: Standard Deviation shows how much variation there is from the average value. Accordingly, 68% of the data will fall within +/- 1 Standard Deviation of the mean and 95% of the data will fall within +/- 2 Standard Deviations of the mean.

As for our results, for RCP8.5, the mean year of climate departure was 2047 (i.e. average year from all 54,000 locations on Earth considered in our study) and the Standard Deviation was +/- 14 years. It follows, that 68% of locations on Earth will have climate departure between 2033 and 2061 and 95% of locations between 2019 and 2075.



Image from: http://en.wikipedia.org/wiki/Standard_deviation