DODO RESULTS Campaign Averages BAe146 Nephelometer Findings Claire

DODO RESULTS: Campaign Averages & BAe-146 Nephelometer Findings Claire Mc. Connell Ellie Highwood Acknowledgements: Paola Formenti, Met Office, FAAM

DODO 1 Feb 2006 DODO Flights DODO 2 Aug 2006

DODO Average Properties DODO 1 ω0550 Observations Mie 0. 98 ± 0. 02 0. 98 0. 96 ± 0. 03 0. 96 Refractive index g 550 σext 550/m-1 kext 550/m 2 g-1 DODO 2 1. 53 - 0. 001 i 1. 53 - 0. 0029 i 0. 69 3. 13 E-03 ± 1. 88 E-03 1. 15 E-04 ± 1. 02 E-04 0. 88 1. 15

Under-sampling in the Nephelometer • Coarse particles are lost in the inlet of the nephelometer • During SHADE (2000) AODs derived from scattering on the C-130 required multiplication of 1. 5 to agree with AERONET AODs, implying an inlet cut-off between 1. 3 and 1. 9μm radius. (Haywood et al. , 2003) • Experiments on BAe-146 nephelometer: 1. The intercomparison between the BAe-146 and the NASA DC-8 during DODO 2 allowed a comparison between the nephelometers 2. Size distributions measured during DODO & Mie code allow an analysis of the under-sampling 3. Under-sampling can be estimated through comparisons with ground-based sites • These processes allow an estimation of the factor of sampling missed and of the cut-off radius on the BAe-146 nephelometer

Intercomparison with NASA DC-8 Uncorrected nephelometer scattering for 3 straight level runs – bold=BAe-146, light=DC-8 • BAe-146 and DC-8 show same spatial variability in scattering – correlation coefficients vary between 0. 93 -0. 96 • Difference in absolute scattering measured – average ratio of DC-8/BAe-146 uncorrected scattering = 2. 1 • Hypothesis: The BAe-146 nephelometer is underestimating the total DODO dust scattering at 550 nm by a factor of 2 -3, and is only measuring submicron (diameter) scattering

Comparison with AERONET • AODs for DODO 2 calculated by integrating BAe-146 profile data • AERONET AODs for the same times are a factor of 3. 2 larger than those from the nephelometer profiles (but factor varies between 2 and 4. 5) • Results consistent with the DC-8 intercomparison Flight Profile Uncorrected AERONET τ550 Correction factor τ440, Level 2. 0 B 236 P 10 0. 11 2 0. 2 B 237 P 8 0. 23 2 0. 5 B 238 P 1 0. 13 4. 5 0. 6 B 238 P 9 0. 15 3. 3 0. 5 B 242 P 1 0. 04 10 0. 4 B 242 P 11/P 12 0. 10 4 0. 4 Average 4. 3 AODs from BAe-146 profiles and from AERONET station at Dakar

Estimation of Nephelometer Cut-off Radius • The size distribution was measured by the PCASP and SEM sizing during DODO • ‘Cut-off test’ – average size distribution successively ‘cut-off’ at decreasing radii to mimic the loss of the coarse mode • Each size distribution run through Mie code to obtain scattering/absorption properties • ‘Cut-off Factor’ (C) calculated – factor required to match the ‘cutoff’ scattering/absorption/extinction with that of the full size distribution Size distributions available for use in the cut-off test

Results from Cut-off Tests • Submicron sampling (r<0. 5μm) results in a scaling factor of 3 for the nephelometer • ω0 is considerably lower if the full size distribution is considered

Conclusions on Cut-off Radius on BAe-146 Nephelometer • Submicron sampling on the 146 neph would result in a cut-off factor of 3 • The AERONET-derived scaling factor suggests a cut-off of 0. 5μm • DC-8 intercomparison-derived scaling factor suggests a cut off of 0. 8 -1μm • Suggested cut-off, around r=0. 5μm, appears to be smaller than experienced for dust during SHADE 2000 Caveats • Mie work assumes spherical particles • Cut-off factors may change with different size distributions (e. g. more/less coarse mode present)

Future Work Observations • Results from radiometers • Results from Edwards & Slingo radiation code • Chemistry – Iron Content, for deposition – Refractive indices, for comparison with derived values • Specific case studies, comparison to remote sensing Modelling • DODO – dust modelling, comparisons to observations, deposition estimates