ESA STSE
Earth Observation Land Data Assimilation Scheme
EO-LDAS Final Presentation Workshop
Frascati, 25th March 2011
EO-LDAS field campaign Gebesee 2010
measurement concept, satellite data and first results
Matthias Forkel1,2, Hans-Jörg Fischer1, Hannes Tünschel1, Sören Hese1 & Christiane Schmullius1
1 Friedrich Schiller University of Jena, Department for Earth Observation 2 now at Max Planck Institute for Biogeochemistry Jena
1 Scope of the field campaign
- Collection of in-situ and satellite data for validation of
the EO-LDAS prototype
- Plan overall field campaigns and data collection
- Generate schedule of expected overpasses
- Liaise with EO suppliers - Undertake fieldwork according to EO data
collection schedule
- Package field data with metadata
2 Gebesee test site
Fig.: Location of the Gebesee test site in the Thuringian basin near Erfurt
2 Gebesee test site
Advantages of the site large fields (~ 1 km²), small villages relatively homogenous topography and soils → appropriate for medium resolution remote sensing applications presence of an eddy flux tower (51.1001°N, 10.9143°E, 162 m a.s.l.)
good contacts to an open-minded farmer's cooperative
2 Gebesee test site
Soil types
relatively homogenous
topography
and soils: mostly loess soils
and in the south clay
clay
loess
3 Measurement concept
Fields and crop types
- winter wheat: field A, E
- rape: fields B, F
- winter barley: fields C, G
- maize: fields D, H
- primary fields A-D: nearly
weekly measurements
- secondary fields E-H:
additional measurements
Fig.: Sample fields
3 Measurement concept
Sampling Units (SU)
up to 4 SUs per field with 5 points to
capture pixel-scale variability
requirements for SU location:
- >100 m distance to field border
- represent different soil types
Fig.: Design of a sampling unit
Fig.: Location of sampling units
3 Measurement concept
4/20 4/23 4/29 5/12 5/14 5/21 6/4 6/8 6/10 6/16 6/24 7/1 7/2 7/7 7/16 7/22 7/30 8/4 8/10
A 3 4 4 2 3 4 3 3 3 3 4 3 3 3 3
B 4 4 2 2 3 3 3 3 3 4
C 4 4 4 1 4 4 3 3 3 3
D 3 3 3 3 3 3 2 2 2 2
E 4 4 4 3 3 2 3 3
F 4 4 3 3
G 4 4 4 3
H 1 3 3 1 3 3
Tab.: Measurement days and SUs per day and field
- 19 measurement days
3 Measurement concept
Variables
Spectra: 1 per point + 1 as transect across SU = 6 per SU
Leaf area index: 4 values per SU
Canopy cover: 1 value per SU
Vegetation height: 1 value per SU
Surface soil moisture: 3 per point = 15 per SU
Atmosphere Optical Thickness: 1 per SU
Spectrometer measurements
- ASD field spec 3
- spectral range 350 – 2500 nm
- 3 detectors
- spectral resolution:
3 nm @ 350-1000 nm
10 nm @ 1000-2500 nm
- resampled to 1 nm channels
- Raw data as ASD binary files
Field measurements:
- optimization and white reference with spectralon
- 30 measurements at each sampling point
- 30 measurements during a transect through the SU
= 180 single measurements for each SU
3 Measurement concept
3 Measurement concept
Leaf Area Index
- Licor LAI 2000
- fisheye lens
- five detectors for different angles
- 4 LAI values for each SU
- at each corner point of the SU
Canopy cover measurements
- Scalebar 60 cm x 200 cm
-- 10 cm raster
- Photographs taken about 1 m above
vegetation
Estimation of Canopy cover:
- Definiens developer algorithm
- Classification of vegetation and soil
- Ratio calculation
- Canopy cover in percentage
3 Measurement concept
Vegetation height
- photograph of a scalebar in
front of the crops
- 3 m distance between
camera and scale
- estimating average height
manually by interpreting the
photographs
3 Measurement concept
Soil moisure measurements
- ThetaProbe ML2X Soil Moisture Sensor
with 1% accuracy
- 3 measurements at each sampling point
3 Measurement concept 3 Measurement concept
Aerosol optical thickness
Microtops 2 sun photometer
- 5 channels (440, 500, 675, 870, 1020 nm)
- 1 measurement at each SU
3 Measurement concept
Soil laboratory test
- soil sample of each SU
- dried to 0% moisture
- spectrometer measurements of each
sample at 5% steps of moisture increase to
saturation
- analysis of each sample according to the
moisture increase steps with ASD view
spec pro software
4 Soil spectra
4 Soil spectra
4 Soil spectra
Fig.: Soil spectra (loess chernozem) depending on moisture contents
Increasing wetness
0%
40%
Landsat 5 08/05/2010
Formosat-2 25/07/2010
Spot 22/05/2010
Rapideye 09/07/2010
10 km
5 Satellite data
Acquired EO data
6 Results
Leaf Area Index
Fig.: Maps of SU-averaged LAI
6 Results
Leaf Area Index
Fig.: Field-averaged cycles of LAI and standard deviation
loess
clay
6 Results
Leaf Area Index – Spatial variability inside SU, field and crop type
Fig.: Coefficient of variation of LAI in wheat, rape, barley and maize
6 Results
Spectra A) winter wheat B) rape
C) winter barley D) maize
Fig.: Spectra 2010-06-16
6 Results
Spectral indices
Fig.: Field-averaged NDVI, NDMI and PRI and standard deviation
NDVI
Win
ter
wh
ea
t
NDMI PRI
Ra
pe
W
inte
r b
arl
ey
Ma
ize
Many thanks to the EO-LDAS team!
Many thanks to the EO-LDAS team!
Andrius Ramanauskas Caroline Baumgart
Eric Thomas Falko Stier
Franziska Behnsen Hannes Tünschel
Hannes Werner Jana Peukert Kerstin Traut
Lisa Hagedorn Lisa Wedekind
Marius Bockwinkel Martin Faber
Martin Lindner
Participiation grouped by study: 8 students from M. Sc. Geoinformatics = 43% 2 M. Sc. Geography 16 B. Sc. Geography = 52% 1 art history 1 political science 1 nutrition science
Martin Thurner Mathias Müller
Max Tobaschuss Miguel Kohling Peggy Bierbaß Peter Schmider Rene Michaelis
Robert Wolff Sebastian Willi Oehmke
Susann Förster Thomas Brockmann Thomas Schiemann
Tillmann Lösche Tino Wunderlich – B.Sc. thesis
Thanks to UCL for providing measurement instruments!
6 Results
Canopy cover
Fig.: Field-averaged cycles of canopy cover and standard deviation
6 Results
Vegetation height
Fig.: Field-averaged cycles of vegetation height and standard deviation
6 Results
Soil moisture
Fig.: Field-averaged cycles of soil moisture and standard deviation
6 Results
AOT
Fig.: Aerosol optical thickness