? Demonstrate/understand spatial heterogeneity of soil erodibility ? Constrain natural sources of dust, increase confidence in identifying anthropogenic emissions Definitions: Soil aeolian erodibility S is the efficiency with which soil produces dust for a
Spatial Heterogeneity in Aeolian Erodibility: Uniform, Topographic, Geomorphic, and Hydrologic Hypotheses Charlie<zender@uci.edu>, David and Omar 1Department of Earth System Science, University of California at Irvine; 2Joint Center for Earth Systems Technology, University of Maryland Baltimore County Objectives: Demonstrate/understand spatial heterogeneity of soil erodibility Constrain natural sources of dust, increase con dence in identifying anthropogenic emissions
1 Zender
1, Newman
2 Torres
Methods: Four three-year DEAD simulations driven by NCEP 1990–1992 meteorology in MATCH CTM[Rasch et al., 1997] (T62/L28, t= 40 min.). Objectively evaluate hypotheses using three independent datasets:
All three spatially varying erodibility We compared model-simulated AOD to the Large inter-model spread (%) occurs for stations hypotheses produce signi cantly better TOMS-observed AAI and AOD in three in and downwind of source regions, including agreement with station and satellite data than important desert regions. All data are gridded to Sal Island (57%) and Cape Point (2500%), the null (Uniform) hypothesis. T62 and points with vegetation V> 0.3 m2 m 2 Kaashidhoo (50%), Jeju (310%) and Okinawa The Geomorphic erodibility hypothesis are removed to avoid contamination by biomass (280%). Surface observations of concentration U. Miami seasonal& annual mean dust performs best overall, but results vary by burning aerosol. and deposition show minimal inter-model spread region and by metric. concentration De nitions: at most non-dusty stations. Annual mean deposition at 11 stations[Ginoux These results support the hypothesis that dust Soil aeolian erodibility S is the ef ciency with et al., 2001] emission“hot spots” exist in regions where which soil produces dust for a given Spatial correlation with TOMS AAI/AOD alluvial sediments have accumulated and may meteorological forcing. S is dimensionless be disturbed[Prospero et al., 2002]. factor appearing in formulation of dust For DEAD description and evaluation of 1990s Our physically-based erodibility hypotheses mobilization mass ux Fd,j[kg m 2 s 1][Zender climatology: help explain dust observations in some et al., 2003] doc.guandang.net ZBN03.pdf regions, particularly East Asia, and can be I used to help discriminate between natural and Mi,j (1) Results: Fd,j= AmSαQs anthropogenic soil emissions. i=1 Annual Concentration (µg m 3) 10 2
Spatial Correlation with Satellite Observations:
Hypothesis Evaluation with Station Observations:
Conclusions:
10
3
Mean Annual Deposition (g m 2 yr 1)
Unity Topo Geo Hydro
5
4
1
10
2
Unity Topo Geo Hydro
10
10
1
10
9
Model
Model
2
11
10
1
3
1
11
10
0
1413
2
4
12
17
7 6 18
8
10
0
19
15
10
1
16
10
1
7
6 3 5
9 8
10
1
10
0
10
1
10
2
10
1
10
0
Observed
10 Observed
1
10
2
10
3
where Am is fraction of ground tha
t is erodible,α[m 1] is sandblasting mass ef ciency, Qs[kg m 1 s 1] is saltation mass ux, and Mi,j is mass fraction of source mode i carried in transport bin j. Visit doc.guandang.net for more details.
Strategy: Compare performance of four differenct S hypotheses in mineral Dust Entrainment And Deposition (DEAD) model Uniform: globally uniform S. Emissions determined by instantaneous local meteorology, vegetation, soil moisture. Topographic: S characterized by relative elevation in basin[Ginoux et al., 2001] Geomorphic: S characterized by upstream area. Hydrologic: S characterized by present day surface runoff
Figure 6: Mean 1980–2001 TOMS AAI and predicted Figure 3: Predicted annual mean mass emissions ux dust AOD for 1990–1992 in the Saharan desert and Saudi[µg m 2 s 1] for (a) Uniform S, (b) Topographic S, (c) Ge- Arabian peninsula, East Asia, and Australia. (a) TOMS omorphic S, (d) Hydrologic S. Scale is logarithmic. AAI, (b) Uniform S, (c) Topographic S, (d) Geomorphic S, (e) Hydrologic S. r is the spatial correlation coef cient between the experiment and TOMS.
Figure 7: Predicted and observed climatological mean surface concentration of dust inµg m 3 at U. Miami stations. Dashed lines indicate factor of two disparity. Site numbers are 1. Barbados, 2. Miami, 3. Bermuda, 4. Izana, 5. Sal Island, 6. Mace Head, 7. King George Island, 8. Cape Point, 9. Kaashidhoo, 10. Jeju, 11. Okinawa, 12. Enewetak, 13. Midway, 14. Oahu, 15. Nauru, 16. American Samoa, 17. New Caledonia, 18. Norfolk Island, 19. Cape Grim. Axes are logarithmic. (b). Predicted and observed climatological mean dust deposition ux in g m 2 yr 1 at 11 stations compiled by Ginoux et al.[2001]. Site numbers are 1. French Alps (45.5N, 6.5E), 2. Spain (41.8N, 2.3E), 3. Midway (28.2N, 177.35W), 4. Miami (25.75N, 80.25W), 5. Oahu (21.3N, 157.6W), 6. Enewetak (11.3N, 162.3E), 7. Fanning (3.9N, 159.3W), 8. Samoa (14.25S, 170.6W), 9. New Zealand (34.5S, 172.75E), 10. Takla Makan (40.0N, 85.0E), 11. Tel Aviv (32.0N, 34.5E). Axes are logarithmic.
Linking erodibility S to regional geomorphology/hydrology helps constrain natural sources of dust, and thus increases con dence in anthropogenic emissions identi ed using residual methods.
Future Work: Examine Hydrologic hypothesis with wet period (mid-Holocene) runoff Use TOMS/MODIS/AMSR-E to determine empirical S, subtract natural S to obtain anthropogenic S Apply to regional forecasts of Columbia Plateau, Aral Sea
Asante, K. O., Approaches to continental scale Table 1: Spatial Correlation of Model AOD with TOMS river ow routing, Ph.D. thesis, The University 2 a N N AAI and AOD 1 1 of xi yiat Austin, Austin, Texas, 2000. Texas 2; (xi yi) RMSrel= RMSabs= N i=1 N i=1 xi Ginoux, P., M. Chin, I. Tegen, J. Prospero, Sahara East Asia Australia (2) B. Holben, O. Dubovik, and S.-J. Lin, Sources Geomorphic Hypothesis best explains station Figure 4: Pr
edicted annual mean dust burden[mg m 2] Hypothesis AAI AOD AAI AOD AAI AOD observations: and distributions of dust aerosols simulated for (a) Uniform S, and change relative to (a) due to (b) ToUniform 0.55 0.18 0.81 0.78 0.12 0.27 Table 3: Annual Mean Concentration at 19 U. Miami pographic S, (c) Geomorphic S, (d) Hydrologic S. Scale is with the GOCART model, J. Geophys. Res., stationsa Topographic 0.69 0.34 0.76 0.70 0.64 0.55 non-linear. 106, 20555–20273, 2001. Geomorphic 0.65 0.37 0.91 0.77 0.57 0.32 Linearb Logarithmicc Jenson, S. K., and J. O. Domingue, Extracting Hypothesis r m b RMSabs RMSrel r m b RMSabs RMSrel Hydrologic 0.59 0.43 0.71 0.67 0.70 0.51 topographic structure from digital elevation Uniform 0.91 2.04 4.01 16.06 0.88 0.84 1.00 0.11 0.52 2.12 a Number N of valid T62 gridpoints is 290, 77, and 59 for Saharan, East Asian, geographic information system Topographic 0.96 1.58 2.30 8.60 0.85 0.78 1.03 0.20 data for 2.80 0.68 and Australian regions, respectively. Highest r for TOMS AAI in each region is Geomorphic 0.94 1.74 2.96 11.28 0.83 0.91 0.93 0.00 analysis, Photogrammetric Engineering and 0.33 1.35 boldfaced. Hydrologic 0.98 1.47 1.97 6.71 0.81 0.81 1.03 0.18 Remote Sensing, 54, 1593–1600, 1988. 0.60 2.46 Table 2 compares the annual mean dust budget of Prospero, J. M., P. Ginoux, O. Torres, and the control to the three experiments. Mass path is S. Nicholson, Environmental characterization prescribed, so lifecycle varies with proximity of sources to regions of wet deposition. Note 15% Table 4: Biases inaSimulation of Annual Mean Deposi- of global sources of atmospheric soil dust tion at 11 stations derived from the NIMBUS7 TOMS absorbing change in emissions and turnover time. Table 2: Climatological Budget Statistics aerosol product, Rev. Geophys., 40, 2–1–2–31, Linearb Logarithmicc Figure 5: Predicted annual mean mass deposition ux Quantity Uniform Topo Geo Hydro Hypothesis r m b RMSabs RMSrel r m b RMSabs RMSrel 2002. 2 s 1] for (a) Uniform S, (b) Topographic S, (c) Ge[µg m Burdena 19.3 19.3 19.3 19.3 Uniform 0.97 0.07 2.06 125.34 0.84 0.93 0.77 0.15Rasch, P. 0.78 N. M. Mahowald, and B. E. Eaton, 0.45 omorphic S, (d) Hydrologic S. Scale is logarithmic. J., Topographic 1.00 0.21 1.35 107.52 0.74 0.95 0.85 0.14 0.37 0.73 Emissionb 1438 1535 1638 1632 Representations of transport, convection, and Geomorphic 1.00 0.42 0.26 78.66 0.66 0.96 0.90 0.14 0.32 0.71 c Lifetime 5.0 4.7 4.4 4.4 Hydrologic 1.00 0.13 0.97 117.44 0.68 0.94 0.78 0.21 the hydrologic cycle in chemical transport 0.45 0.78 a models: Implications for the modeling of Value of best metric in each category is boldfaced. Globalτ d 0.033 0.034 0.034 0.034 b Linear correlation coef cient r and best t parameters to y= mx+ b where x are short-lived and soluble species, J. Geophys. a Prescribed atmospheric burden in Tg observed data and y are simulated. c b Statistics computed using log x and log y in place of x
and y. Emissions of D< 10µm particles in Tg yr 1 Res., 102, 28127–28138, 1997. c Turnover time in days d
Spatial correlations show that that regional topographic-geomorphic-hydrologic features explain observed distributions better than Uniform hypothesis. Value of best metric in each category is boldfaced.
Linear statistics preferentially weight dusty stations, so we compute logarithmic statistics too. Absolute and relative RMS biases RMSabs and RMSrel, respectively, computed using
References
Figure 1: Distinct geomorphic basins are determined by back lling DEM algorithm[Jenson and Domingue, 1988], modi ed to preserve true internal drainage basins[Asante, 2000].
Value of best metric in each category is boldfaced. b Linear correlation coef cient r and best t parameters to y= mx+ b where x are observed data and y are simulated. RMS de ned in (2) and (2). c Statistics computed using log x and log y in place of x and y.
a
Optical depth at 0.63µm
Zender, C. S., H. Bian, and D. Newman, The

热门文档