progress and plans for modis validation and new algorithm ......jan-peter muller, university college...
TRANSCRIPT
DEPARTMENT OF GEOMATIC ENGINEERING
Progress and Plans for MODIS validation and new algorithm developmentJan-Peter Muller, University College LondonProfessor of Image Understanding and Remote Sensing
MODIS & MISR Science Team Member (NASA EOS Project), HRSC (ESA Mars Express) CoIwith contributions from Catherine Naud, SJ Chan, Chris Doll, Jung-Rack Kim, Zhenhong Li, Mercedes Sole-Chomorro (UCL)
Objectives for next 3 yearsDevelop automated algorithms for the extraction of geophysical parameters from
MODIS using geomatic engineering, image understanding (IU) techniques and 3D radiative transfer theory
Develop extensions of existing MOD43 global BRDF/Albedo product to take topography into explicit account
Develop new products for the climate modelling community based on fusion of MODIS and SRTM near-global topography
Develop fused MODIS-MISR products to exploit the complementarity between their spectral and angular aspects (clouds, aerosols, surface BRDF/albedo)
Develop fused MODIS-DMSP products to explore the anthrogenic components of global change
Develop global validation strategies using independent data-sets as well as inter-comparison between different sensors on the same EO platform.
Publish individual validation case studies on the web to provide a detailed reference for product users.
Validation of Cloud and Water Vapour Properties from the NASA Terra MODIS and MISR instruments (EU-CLOUDMAP2)
Long time series inter-comparisons of cloud-top heights and cloud fractions from ground-based active and passive instruments at ARM-SGP, OK and Chilbolton Radar Observatory, UK
Good agreement between thermal IR all-sky camera and MODIS cloud fractions
All cases per cloud situation: good agreement between MISR, radar and MERIS for single level (non-broken) clouds (#: large box indicates high cloud whilst smallerboxes indicate mid-level cloud)
Validation of Aerosol Properties from the NASA Terra MODIS and MISR instruments with street-level urban pollution
A E R O S O L D A T A
U RBA N A ER OS OL D A T ASATELLITE AEROSOL RS DATA
Cloud-free atmosphere retrievals
MODISAEROSOLPRODUCT
MISRAEROSOLPRODUCT
L.A PM-10 & PM-2.5 (µg/m3)
London PM-10 & PM-2.5 (µg/m3).
GROUND BASED AERONET DATA
AOT Validation
30-arc-second Global Population Dataset(LANDSCAN)
30-arc-second Global elevation data(GTOPO30)
AOT
Ångströmexponent
AOT
R E S U L T S: Inter-comparison MISR /MODIS AOT vs. PM-2.5 for L.A.& London
0
0 .05
0.1
0 .15
0.2
0 .25
0.3
0 .35
0 1 0 2 0 3 0 4 0 5 0 6 0 7 0
M
OD
IS A
OT
0
0.1
0.2
0.3
0.4
0.5
0.6
MODIS AOT
MISR AOT
M
ISR
AO
T
13-11-0014-2.01
24-1-0129-5-0113-2-0130-5-01
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
0 5 10 15 20 25 30 35 40 45
MO
DIS
AO
T
16-3-01
17-4-01
3-4-01
0
0.20.40.6
0.81
1.21.41.6
1.82
0 10 20 30 40 50
PM-2.5 24-h aver.
MIS
R A
OTM
ISR
AO
T
MIS
R A
OT
MO
DIS A
OT
PM-2.5 24-hour aver.
PM-2.5 24-hour aver.
PM-2.5 24-hour aver.
Los Angeles, 25th Nov. 2000
London, different dates
• MISR seems to correlate better with PM2.5 than MODIS
• Larger number of stations would give less uncertainty on the results
R E S U L T S: Inter-comparison MODIS/MISR AOT vs. PM-10 and PM-2.5 for L.A.
13th Nov. 2000
NorthSea
London
English Channel
17th April 2001
NorthSea
London
English Channel
MODIS AOT
MISR AOT
Locations of the monitoring stations within the London urban area
•Diff icult to obtain total coverage over the London urban area due to cloud cover
• PM10 and PM2.5 is given in 24h and hourly averages
• PM2.5 is reported at only two stations in central London:Marylebone and Bloomsbury
R E S U L T S: Inter-comparison MISR /MODIS AOT vs. Global elevation dataset
Hill-shaded DEM (GTOPO30), California (U.S.A) MODIS AOT 17 th Jan 2001 MISR AOT 4th Feb 2001
• Good visual correlation between AOT retrievals and topography for both California
• Strong temperature inversions in the L.A valley is one of the causes of long, low visibility episodes and higher health risks
• San Joaquin Valley is trapped between the Sierra Nevada Mountains and the coastal range where pollution concentrations are high
Land Surface Albedo Properties from MODIS instrument Anthropogenic activities from the DMSP OLS instrument
Night-time Lights as an indicator of human activity
Excellent at identifying human activities in developed parts of the world Also highlights human activity at sea such
as shipping fleets and off-shore gas fields. Clear economic component when analysing the data
Mapping CO2 from night-time lightsMODIS 16-day broadband albedo at 1km over Europe Anthropogenic signatures in MODIS Albedo
Multi-spectral image 1 m resolution DEM
NDVI 1 m pan image
Example Application of IU techniques to Mars
Reliable tree and building detection using the fusion of multi-spectral and Digital Surface Model information and both supervised and unsupervised classification. 3D radiative transfer can then be applied to the 3D landscape objects to understand better the spectral and angular signatures seen by MODIS
Automated crater detection in MGS-MOC images
Automated building detection from 1m IKONOS
Such automated high-level feature extraction based on image understanding techniques will be applied to MODIS products to try to extract climate change signatures
-0.20
-0.16
-0.12
-0.08
-0.04
-0
Albedo
Mapping economic activityfrom night-time lights
MERIS and Radar CTH (km) MISR, MODIS and Radar CTH (km) Cloud situation Date and location
Time of overpass (UT)
MERIS Radar Time of overpass (UT)
MISR MODIS Radar
Low continuous 2003-02-15, SGP
16:58 1.9 2.0 17:30 2.2 0.8 2.1
2003-05-11, CFARR
10:42 CTHs too variable 11:10 1.0 0.6 1.1 Low broken
2003-06-12, CFARR
10:35 1.7 1.6 11:10 1.5 0.9 2.0
Low multi-layer 2003-05-27, CFARR
10:37 1.2 2.5/1.4* 11:10 1.2 2.3 2.5/1.4*
Mid-level continuous
2003-06-07, SGP
16:39 4.3 4.6 17:30 4.5 6.1 4.7
Mid-level broken
2003-06-09, SGP
17:15 4.9 6.2 17:15 5.4 7.1 6.2
2003-05-15, SGP
17:01 5.0 11.7 17:25 9.0 10.6 11.6 High multi-layer
2003-05-31, SGP
16:58 5.5 12.4 17:25 9.0/5.3# 10.6 11.8
DATE Time(UT)
ASTIC CF(%)/manual
ASTICimage
ASTIC Cloudmask
Comments
19 June2003
11:14:06 60.8/100
26 June2003
11:12:35 86.0
03 July2003
11:22:06 80.3
05 July2003
11:14:06 61.1/100
06 July2003
11:51:06 71.8/92
07 July2003
11:56:06 72.9
08 July2003
11:44:06 58.1/100
Excellent agreement between MOD05 and GPS Precipitable Water Vapour
Barnsley, M.J., P.D. Hobson, A.H. Hyman, W. Lucht, J.P. Muller, and A.H. Strahler, Characterizing the spatial variability of broadband albedo in a semidesert environment for MODIS validation, Remote Sensing of Environment, 74 (1), 58-68, 2000.
Doll, C.N.H., J.P. Muller, and C.D. Elvidge, Night-time imagery as a tool for global mapping of socioeconomic parameters and greenhouse gas emissions, Ambio, 29 (3), 157-162, 2000.
Li, Z.H., J.P. Muller, and P. Cross, Comparison of precipitable water vapor derived from radiosonde, GPS, and Moderate-Resolution Imaging Spectroradiometer measurements, Journal of Geophysical Research-Atmospheres, 108 (D20), art. no.-4651, 2003.
Muller, J.P., A. Mandanayake, C. Moroney, R. Davies, D.J. Diner, and S. Paradise, MISR stereoscopic image matchers: Techniques and results, Ieee Transactions on Geoscience and Remote Sensing, 40 (7), 1547-1559, 2002.
Naud, C., J.P. Muller, and E.E. Clothiaux, Comparison of cloud top heights derived from MISR stereo and MODIS CO2-slicing, Geophysical Research Letters, 29 (16), art. no.-1795, 2002.
Naud, C., J.P. Muller, M. Haeffelin, Y. Morille, and A. Delaval, Assessment of MISR and MODIS cloud top heights through intercomparison with a back-scattering lidar at SIRTA, Geophysical Research Letters, 31 (4), art. no.-L04114, 2004.
Sample publications (2000-2004)