4 An Overview of the Changing Climate in Northern Alaska
The study employs temperature, solar radiation, and CO2data reported by the Barrow (BRW) Atmospheric Observatory. This is one
of the baseline observatories of the Earth System Research Laboratory
(ESRL), Global Monitoring Division (GMD), of the National Oceanic and
Atmospheric Administration (NOAA). It is located near sea level about 8
km east of Utqiaġvik (formerly Barrow), Alaska at 71.3230 degrees north
and 256.6114 degrees West (Vasel et al., 2020). Continuous atmospheric
measurements of CO2 have been recorded at this
observatory since July 1973 (Thoning et al., 2021). Herbert et al.
(1986) discuss how the data are processed. Peterson et al. (1986)
discuss the first ten years (1973-1982) of operations and report
consistency of the Barrow results with the reported data from four
neighboring locations. Tans and Thoning (2020) provide a general
overview of the methods used to collect and process the
CO2 data at Mauna Loa, one of NOAA’s other baseline
observatories. Along with the hourly temperature data corresponding to
BRW, the CO2 data for BRW were downloaded using the
following link: (http://www.esrl.noaa.gov/gmd/dv/data/ ).
Measurements of downward total solar irradiance have been reported at
the BRW observatory since January 1976. Before 1998, the data were
reported at three minutes intervals. The data were subsequently reported
at one-minute intervals. For this study, the reported values were rolled
up to hourly averages. Data were dropped from the analysis if the number
of valid minutes of data for an hour was less than 15.
Consideration was given to the inclusion of CH4 data in
the analysis. This action would have resulted in the loss of 26,381
hourly observations due to unavailable or invalid CH4measurements. (the collection of the CH4 data commenced
in 1986 but was subsequently suspended for about nine months in
2012/2013). The probable effect of this data loss on model convergence
was an important consideration in excluding this variable from the
analysis, model convergence being one of the major challenges of the
methodology employed in this paper (STATA, 2021, p. 33). The omission of
CH4 and other variables reflecting greenhouse gas
concentrations represents a shortcoming in the analysis.
The sample for this study spans from 1 Jan 1985 through 31 Dec 2015.
Data before 1 Jan 1985 were not employed in this study because the
reported downward total solar irradiance data largely did not meet
ESRL’s standards before that date. For example, only about 31% of the
downward total solar irradiance values in 1984 were deemed by ESRL to be
valid. The 1 Jan 2016 - 31 Aug 2017 time interval is reserved for
out-of-sample analysis. The evaluation period terminates on 31 Aug 2017
because of a significant data availability issue.
In thinking about meteorological issues at BRW, it is useful to begin by
first noting the extremes and high level of variability in the level of
downward total solar irradiance at this location. In terms of
variability, the data from 2014 is instructive (Figure 2). Concerning
the extremes, there are about 67 days of virtually total darkness each
year ( about 18 Nov to 22 Jan), while the sun does not completely set
from 11 May to 31 Jul.