Aj. Krueger et al., VOLCANIC SULFUR-DIOXIDE MEASUREMENTS FROM THE TOTAL OZONE MAPPING SPECTROMETER INSTRUMENTS, JOURNAL OF GEOPHYSICAL RESEARCH-ATMOSPHERES, 100(D7), 1995, pp. 14057-14076
The total ozone mapping spectrometer (TOMS), first flown on the Nimbus
7 satellite, has delivered an unanticipated set of unique information
about volcanic plumes because of its contiguous spatial mapping and u
se of UV wavelengths. The accuracies of TOMS sulfur dioxide retrievals
, volcanic plume masses, and eruption totals under low-latitude condit
ions are evaluated using radiative transfer simulations and error anal
ysis. The retrieval algorithm is a simultaneous solution of the absorp
tion optical depth equations including ozone and sulfur dioxide at the
four shortest TOMS wavelengths and an empirical correction based on b
ackground condition residuals. The retrieval algorithm reproduces mode
l stratospheric sulfur dioxide plume amounts within +/-10% over most c
entral scan angles and moderate solar zenith angles if no aerosols or
ash are present. The errors grow to 30% under large solar zenith angle
conditions. Volcanic ash and sulfate aerosols in the plume in moderat
e optical depths (0.3) produce an overestimation of the sulfur dioxide
by 15-25% depending on particle size and composition. Retrievals of t
ropospheric volcanic plumes are affected by the reflectivity of the un
derlying surface or clouds. The precision of individual TOMS SO2 sound
ings is limited by data quantization to +/-6 Dobson units. The accurac
y is independent of most instrument calibration errors but depends lin
early on relative SO2 absorption cross-section errors at the TOMS wave
lengths. Volcanic plume mass estimates are dependent on correction of
background offsets integrated over the plume area. The errors vary wit
h plume mass and area, thus are highly individual. In general, they ar
e least for moderate size, compact plumes. Estimates of the total mass
of explosively erupted sulfur dioxide depend on extrapolation of a se
ries of daily plume masses backward to the time of the eruption. Error
s of 15-30% are ndt unusual. Effusive eruption total mass estimates ar
e more uncertain due to difficulties in separating new from old sulfur
dioxide in daily observations.