Comparison of the aerosol scattering and absorption coefficients measured at KCO and on the C130 aircraft during flybys

Anne Jefferson, Pat Sheridan and John Ogren

NOAA/CMDL Aerosol Group

Contact: ajefferson@cmdl.noaa.gov

NOAA/CMDL measurements of aerosol scattering coefficients on board the C130 aircraft and at Kaashidhoo (KCO) had similar instrument setups. Both platforms included measurements of aerosol scattering and hemispheric back scattering coefficients at three wavelengths using a TSI nephelometer (model 3563). For brevity we will provide only an analysis of the submicron total scattering channel at 550 nm.

There were 11 flybys of the C130 past KCO. Three of these flights had only the sub10 micron size range, days 59, 77 and 80. Day 59 had only 5 minutes of sub10 on the C130. Day 77 and 80 had only sub10 at KCO (no impactor switching), with 3 minutes of C130 intercomparison time in the size range on Day 80 and no sub 10-micron data for Day 77. Due to the shortness of the intercomparison time and large uncertainty associated with this cut size, these three flights have been removed from the analysis. This leaves 8 research flights to intercompare the submicron scattering between the two platforms.

Data in Figure 1 are presented as the mean scattering values plus/minus the associated measurement uncertainty. The KCO data were averaged for 90 minutes before and after the time the aircraft passed the island. The data average time was estimated from the C130 time spent on the low altitude leg, the C130 flight speed and the ambient wind speed. The calculated time for the air mass to pass KCO using these parameters usually extended longer than 3 hours. The time was cut to 3 hours to avoid problems associated with changes in the wind direction at KCO and to have a uniform average time of 3 minutes and thus similar number of data points for all flights. The amount of data from the C130 depended on the low altitude flight leg past the island, the time spent in measuring the two aerosol size ranges and the extent of heater edits (http://www.cmdl.noaa.gov/aero/net/c130/indoex/problem_summary/index.html). These times ranged from 4.8 to 14.2 minutes for the eight flybys.

Associated measurement uncertainty:

Calculation of the measurements uncertainty of each instrument followed a similar protocol as that described in Anderson and Ogren (Aerosol Sci and Tech, 29, 57-69, 1998). The mean values of the green scattering coefficient with the calculated uncertainties are displayed in the above figure. The 95% confidence intervals of the averaged data (derived from the standard deviation of the data) are smaller than the uncertainty error bars denoted in the above figure. The measurement uncertainty associated with the TSI nephelometer was calculated from four known sources:

  1. the precision uncertainty pertaining to instrument noise, drift and an offset between instrument calibrations (2.8 to 3.4 Mm-1)
  2. uncertainty in the instrument temperature and pressure associated with corrections of the data to STP (0.1 to 0.6 Mm-1)
  3. uncertainty in the "truncation" correction which accounts for the nephelometer blocking of near forward scattered light (0.6 to 1.5 Mm-1)
  4. uncertainty in the instrument calibration. This value was calculated from the average percent error of several years of weekly nephelometer CO2 span checks from three CMDL surface measurement sites. This is the percentage error in the measured scattering for CO2, compared to the expected value at the measured temperature and pressure. (1.5 to 3.4 Mm-1)
Uncertainty associated with differences in the aerosol inlets and tubing is expected to be insignificant for submicron aerosol.

Flight Notes:

This discussion includes flights during which the scattering data from the two platforms fell outside the measurement uncertainty of the other. Of the eight C130 flybys of KCO five intercomparison times had submicron aerosol green scattering values that fell well outside the uncertainty of the other platform. The differences between the two measurements on these flights stem from several factors.

Day 51

The C130 green Bsp value exceeds that of KCO by about 30 Mm-1 or 39%. The air mass passing the island was unusual. Within an hour of the aircraft passing the island, there was an abrupt change in wind direction from the NE to the NW. Back trajectory calculations (www.cmdl.noaa.gov/ozwv/traj/plots/kco.html) show that the air mass arriving at 50m originated from the BOB, passed over the southern tip of India and into the AS before passing KCO. The air at the island was a mixture of relatively clean air from the AS and polluted air from the BOB. Scattering coefficients at KCO exhibited an abrupt decline at 51.5, an hour after the flyby, from ~76 Mm-1 down to ~10 Mm-1. On this flyby the aircraft and island may have sampled quite different air masses as the wind direction and Bsp signals were highly variable.

Days 49 and 58

On these days the C130 green Bsp signal exceeded that of KCO by ~ 14% or 14 Mm-1 on day 49 and ~ 32% or 19 Mm-1 on day 58. The reason for the higher C130 scattering values on these days is unresolved. Both instruments had similar RH values and temperatures. The wind direction at the island during both flybys was from the NE at 33-45 degrees. On day 49 the aircraft flew north of the island averting an island plume. On day 58 the aircraft flew SE of the island. Wind speed at this time was ~3.8 m/s, strong enough for the wind to maintain its course. The aircraft likely averted the island plume on this day as well. The Bsp signals were relatively constant over the intercomparison times.

Day 68

The mean C130 Bsp green signal is ~ 42% or 28.6 Mm-1 higher than the signal at KCO. On this particular flight there were extensive heater edits of the C130 data with only 4.8 minutes of C130 data to compare to KCO. The Bsp signal as well as that of the cold CN significantly increased over the course of the flyby. Because of the short amount of C130 Bsp data and an apparent change in aerosol concentration during the flyby, the scattering data between the two platforms did not compare well.

Day 70

During this flyby the C130 green Bsp value exceeded that at KCO by 40% or 27.6 Mm-1. Less than an hour after the flyby (70.54) there was an abrupt change in wind direction from NW to NE, which brought in clean air from the AS, a rapid rise in RH and rain. As a result of these changes the two platforms likely sampled different air masses.

Absorption signal:

An intercomparison of the submicron absorption data shows that the data from the two platforms agree (fall within the uncertainty of the other) for all but one flight, Day 56, where the mean C130 Bap exceeded that of KCO by about 13 Mm-1. For Day 56 the PSAP transmission at KCO over the intercomparison time ranged from 73% down to 48%. The data during this time were on the lower end of the saw tooth curve and such were diminished by the low transmission. The ratio of mean C130/KCO Bap values averaged for all eight flights is 1.09. The uncertainty range shown in the figure below was calculated from the parameters described in Bond et al. (Aerosol Sci. and Tech, 30, 582-600,1999) where the uncertainty is a factor of the instrument precision, noise, flow rate, filter spot size, aerosol green scattering coefficient and correction factors K1 and K2.