MONITORING OF DUST EMISSION IN EASTERN AUSTRALIA: Why two methods are better than one John Leys, Stephan Heidenreich, Tadhg O Loingsigh, Craig Strong and Grant McTainsh 1
Dust is an indicator of landscape health Wind erosion emits dust in to the atmosphere Dust is an approved resource indicator in Australia Dust Storm Index has been the measure since 1998 (McTainsh 1998) Erosion and Landscape Evolution,Alaska,19 Sept 2011 2
Dust for as an indicator of soil condition Poor management practice Low ground cover Condition Report Best management practice Good ground cover Dust source No dust Declining resource condition Improve / maintain condition Erosion and Landscape Evolution,Alaska,19 Sept 2011 3
The problem gaps in BoM data Number of sites being monitored by BoM is declining (810 in 1974 to 424 in 2010) The number of obs /day is also dropping Erosion and Landscape Evolution,Alaska,19 Sept 2011 4
The challenge Can the new high resolution DustWatch record be used to continue on the Dust Storm Index? Erosion and Landscape Evolution,Alaska,19 Sept 2011 5
DustWatch data Continuous monitoring with DustTrak PM 10 15 min readings then 1 min when PM 10 > 25μg/m 3 1 hour averages Sometimes co-located at BoM stations Erosion and Landscape Evolution,Alaska,19 Sept 2011 6
Methods annual (July to June) Pick 3 good BoM sites with co-located DustWatch nodes Calculate the DSI from DustWatch PM 10 data PM 10 concentration visibility BoM weather code DSI DW Compare annual DSI DW vs. DSI BoM Erosion and Landscape Evolution,Alaska,19 Sept 2011 7
Mildura Erosion and Landscape Evolution,Alaska,19 Sept 2011 8
Visibility to PM 10 relationship 10.000 Buronga DustTrak PM10 MaxValue [mg/m3] 1.000 0.100 0.010 y = 0.51x -1.02 R 2 = 0.54 0.001 0.1 1 10 100 Mildura Airport Vis [km] Erosion and Landscape Evolution,Alaska,19 Sept 2011 9
Annual DSI DW vs. DSI BoM 20 1:1 BoM DSI 15 10 5 2005/06 2009/10 2009/10 anomalies Mildura Cobar Moree 01:01 Linear (Mildura) Linear (Cobar) Linear (Moree) 2007/08 2006/07 2010/11 2008/09 2010/11 2008/09 0 2008/09 2010/11 2007/08 0 5 10 15 20 2009/10 DustWatch DSI 10
Mildura 2005/06 The BoM DSI 6.9 driven by a single past weather SDS code on 14/07/2005 BUT: - the TSP in Buronga was very low (5.89 μg/m 3 ) - the visibility records for Mildura showed no reduction in visibility ( i.e. < 30 km) that day. - humidity was too low to justify fog (<80%) - wind speeds were low (~10km/h) The DustWatch data for early on 14/07/2005 shows aerosol in the air but the DustWatch QA suggests it was smoke. Therefore this observation was removed from BoM DSI Erosion and Landscape Evolution,Alaska,19 Sept 2011 11
In 2007/08 missed obs also occurred Cobar 2009/10 The DW DSI is driven by several SDS in September 2009 (DSI = 15.0 out of 21.15) 6 DustWatch Obs every hour. Blue arrows used in DW DSI calc 5 4 3 BoM Obs 2 1 0 DWN concentration (mg/m 3 ) 26 12:00 26 06:00 26 00:00 25 18:00 25 12:00 25 06:00 25 00:00 24 18:00 24 12:00 24 06:00 24 00:00 23 18:00 23 12:00 23 06:00 23 00:00 22 18:00 22 12:00 September 2009 Erosion and Landscape Evolution,Alaska,19 Sept 2011 12
Cobar 2008/09 The DW DSI for 2008/09 is driven by September 2008 (1.15) and March 2009 (2.15). The differences between the BoM and DustWatch arise from the BoM use of the Local Dust Event code (LDE) which has no visibility associated with it AND the code thresholds Date BoM Code PM 10 (mg/m 3 ) DW threshold (mg/m 3 ) 15 Sept 2008 LDE 0.253 MDS > 0.240 03 March 2009 Nothing recorded 0.310 MDS > 0.240 04 March 2009 LDE 0.605 MDS > 0.240 Erosion and Landscape Evolution,Alaska,19 Sept 2011 13
Corrected DSI values 25 But the QA with the DustWatch 20 data is the key!! BoMDSI 15 10 5 y = 0.81x R 2 = 0.99 Current BoM DSI under 1:1 estimates DW DSI by 19% Coincidently there were 19% obs missing of a possible 8760 three hr BoM obs Therefore there is a good relationship between DW and BoM obs after QA 0 0 5 10 15 20 25 DW DSI Erosion and Landscape Evolution,Alaska,19 Sept 2011 14
Conclusions Two methods are better than one: Erosion and Landscape Evolution,Alaska,19 Sept 2011 - Can calculate DSI from DW data - Longevity of monitoring and increased QA DustWatch can increase the QA of BoM records and visa versa BUT we would need a DustWatch site at each BoM site Not possible to extrapolate DW data to neighbouring BoM stations due to temporal variability 15
Questions? JADE DustWatch site Erosion and Landscape Evolution,Alaska,19 Sept 2011 16