Scheduled Downtime
On Wednesday 09 March 2022 from 6am to 10am MT, the website will be down for maintenance

Data from Fig. 2 of Kay et al., BAMS 96, 1333 (2015)


Ralph Young
New Member

I have a question and a request related to the 2015 paper describing the CESM-LE Project [1].

The question: The Monte-Carlo method that was used to model cloud variability [2,3] presumably requires a (pseudo)random number generator. Was the same random-number seed used for each member of the ensemble? Or, did each member use a different sequence of random numbers?

The request: If the curves shown in Fig. 2 of the 2015 paper [1, attached] are available in a format that a mere mortal can handle (txt, dat, csv, xls,...), I would dearly like to have a copy, ideally including the additional 10 members of the ensemble that were computed subsequently. I am hoping that the file or files that were used to construct the figure are already in such a format. I have obtained corresponding curves for the mid-to-upper troposphere [4], and I'd like to compare them with the surface-temperature data in the paper.

Best regards,

[1] J. E. Kay et al., "The Community Earth System Model (CESM) Large Ensemble Project", BAMS 96, 1333 (2015)
[2] R. B. Neale et al., NCAR Technical Note (November, 2012)
[3] R. Pincus et al., J. Geophys. Res. 108 (D13), 4376 (2003)
[4] B. D. Santer et al., Proc. Nat. Acad. Sci. 116, 19821 (2019)


  • Kay BAMS (2015) Fig. 2.jpg
    Kay BAMS (2015) Fig. 2.jpg
    64 KB · Views: 4


Adam Phillips
CVCWG Liaison
Staff member
Hi Ralph,
I created figure 2 of the 2015 Kay et al paper, and just became aware of your August post. My apologies for the delay. The timeseries for figure 2 were not saved, so I had to update my script to analyze all 40 members and to output the timeseries. I have attached the timeseries here as a txt file, along with the updated plot showing all 40 members.
The data is ordered as follows:
Every column = a year, starting in 1449 and ending in 2100. (Note #1: Technically the control run data starts in year 1 and continues to year 401, but for the creation of the plot I created a single array that encompassed all the data. Note #2: The plot shows years 1750-2101.)
Missing value = -99.
Row 1 = Observational values (1920-2013, shown in plot as red line)
Row 2 = Control run values (1449-1849, blue line)
Row 3 = 1st member of CESM1 Large Ensemble (1850-2100, black line)
Rows 4-42 = Members 2-40 of CESM1 Large Ensemble (1920-2100, gray lines)

Hope this helps!


  • fig2.png
    134 KB · Views: 1
  • Kayetal_2015.fig2.timeseries.txt
    187.2 KB · Views: 3