The seasonal cycle of surface albedo of sea ice in the Arctic is estim
ated from measurements made with the Advanced Very High Resolution Rad
iometer (AVHRR) on the polar-orbiting satellites NOAA-IO and NOAA-11.
The albedos of 145 200-km-square cells are analyzed. The cells are fro
m March through September 1989 and include only those for which the su
n is more than 10 degrees above the horizon. Cloud masking is performe
d manually. Corrections are applied for instrument calibration, noniso
tropic reflection, atmospheric interference, narrowband to broadband c
onversion, and normalization to a common solar zenith angle. The estim
ated albedos are relative, with the instrument gain set to give an alb
edo of 0.80 for ice floes in March and April. The mean values for the
cloud-free portions of individual cells range from 0.18 to 0.91. Month
ly averages of cells in the central Arctic range from 0.76 in April to
0.47 in August. The monthly averages of the within-cell standard devi
ations in the central Arctic are 0.04 in April and 0.06 in September.
The surface albedo and surface temperature are correlated most strongl
y in March (R = -0.77) with little correlation in the summer. The mont
hly average lead fraction is determined from the mean potential open w
ater, a scaled representation of the temperature or albedo between 0.0
(for ice) and 1.0 (for water); in the central Arctic it rises from an
average 0.025 in the spring to 0.06 in September. Sparse data on aero
sols, ozone, and water vapor in the atmospheric column contribute unce
rtainties to instantaneous, area-average albedos of 0.13, 0.04, and 0.
08. Uncertainties in monthly average albedos are not this large. Conte
mporaneous estimation of these variables could reduce the uncertainty
in the estimated albedo considerably. The poor calibration of AVHRR ch
annels 1 and 2 is another large impediment to making accurate albedo e
stimates.