I am new to statistics but scraped data from NOAA's web site to plot CO2 level vs time. Is there a good statistical technique to extrapolate past limit of data X axis to near future like 2050? One could cont the curve I guess in a polynomial function. This is a super relevant question as climate change is dangerous.
The NOAA site has the Moana Loa observatory atmospheric CO2 data directly downloadable as text at ftp://aftp.cmdl.noaa.gov/products/trends/co2/co2_mm_mlo.txt - no scraping needed. After removing a few "-99.99" values for "average CO2" that are stated to represent missing data I was easily able to perform an equation search on just over 350 known, named equations with three or less fitted parameters. The search purpose was to find an equation that both fit well and should also extrapolate well due to simplicity. My equation search turned up the Standard Vapor Pressure equation "CO2 = exp(a + (b/year) + c*log(year))" as an excellent candidate, with parameters a = -1.9342529555687474E+03, b = 4.3355326963670307E+05 and c = 2.2673459584210687E+02 yielding RMSE = 2.216 and R-squared = 0.993, here is error plot of this fit with extrapolation of this equation to the year 2040.