Description of data sets collected
1. Digital multispectral air photography
The Ordnance Survey collected aerial photographs of the study area on 9th June 2006 using an Integraph Digital Multispectral Camera. These data have a nominal ground resolution of 25cm and are available in blue, green, red and near infra-red bands together with the necessary metadata to facilitate orthorectification using photogrammetric software.
2. SPOT HRG data
A SPOT-5 HRG half-scene covering the study area and the surrounding countryside was acquired on the 10th June 2006 (scene 029/246). These data include green, red, near infra-red and short-wave infra-red (SWIR) bands, the first three having a nominal ground resolution of 10m (20m for the SWIR). Access to these data is limited because of licensing restrictions, so prospective users should contact the NCAVEO office in the first instance (ncaveo@soton.ac.uk).
3. Remotely sensed data collected on 17th June 2006
On the 17th June, the group acquired satellite data from UK-DMC and CHRIS-Proba, together with airborne data from several airborne imaging spectrometers as well as airborne LiDAR and extensive ground spectral data (Figure 1).
Figure 1. Timeline for the 'golden day', 17th June 2006
Sky conditions on this day were very good for the first period of measurements, but cumulus clouds developed during the late morning as shown by Figure 2.
Figure 2. Radiance of a Spectralon panel measured in a band centred on 500nm during the measurements made on 17th June 2006.
The UK-DMC image was not affected by cloud, but the CHRIS-Proba images acquired later in the morning were badly affected by cloud and cloud shadows. Valid data were acquired from all five CHRIS view angles (Table 2). The UK-DMC data consist of three multispectral bands (green, red, near infra-red) with a nominal ground resolution of 32 m. The study area forms a very small part of the whole DMC image, which has a swath width of over 640 km.
Two aircraft operated over the test area on 17th June. The first was a Cessna 404 operated by the UK Environment Agency (EA) carrying an Itres Instruments Compact Airborne Spectrographic Imager (CASI-3), an Optech 2033 LiDAR and a digital multispectral camera. All three sensors were integrated with the on-board navigation system to provide accurate geometric correction through post-processing. The CASI-3 was operated in hyperspectral mode, with 32 spectral bands and a nominal ground resolution of 1m. The LiDAR was operated at 33 kHz to give a nominal pulse spacing around 1-2 m, and intensity as well as first and last pulse returns were recorded. The digital camera produced colour images 4080 x 4080 in size, with a nominal ground resolution around 40 cm.
Once the Cessna 404 had completed its survey, the second aircfraft, a Dornier 228 operated by the NERC Airborne Research and Survey Facility (ARSF) started collecting data. This had a CASI-2 system on-board, together with an Aisa Eagle imaging spectrometer which acquired data in 244 visible and near IR bands. In each case, the aircraft flew eight adjacent roughly N-S flightlines to cover the whole study area, and then one E-W across the study area, passing directly over the CFARR site (Figure 3).
4. Ground spectral data
Several groups brought field spectroradiometers to the site, and central co-ordination of this activity was provided by staff from the NERC Field Spectroscopy Facility (FSF) who, together with staff from the National Physical Laboratory (NPL), established a calibration laboratory on-site. Several 6m x 6m plastic sheets of known reflectance were placed on the ground within the CFARR site for use as validation targets with the airborne instruments (Figure 4). In addition, an area of asphalt approximately 10m x 10m and another of concrete 40m x 40m in size within the CFARR perimeter were designated as persistent validation targets, and frequent reflectance measurements made of these.
Figure 3. The CFARR site showing the concrete (A), asphalt (B) and artificial targets (C). The image is a subset from one of the CASI-3 flightlines © Environment Agency, 2006.
Three ASD FieldSpec Pro spectroradiometers were used in single-beam, hand-held mode to measure the hemispherical-conical reflectance factor of the crops and the validation targets at various times on the 17th June as well as on the days around the 'golden day'. A dual-beam GER1500 spectroradiometer was used from a mobile trolley to measure the HCRF of the asphalt validation target throughout the period of data acquisition on 17th June. Several Ocean Optics USB2000 spectroradiometers were deployed around the site. One was used with a cosine-corrected receptor to measure the spectral irradiance (400-1100nm) every 30 seconds throughtout the period of measurements on 17th June. Another was used in a manually-operated goniometer to measure the spectral irradiance distribution over the same time period, and another system was used in a dual-beam configuration to measure the HCRF of the artificial validation targets. Finally, a seven-band filter-based spectral radiometer was used to measure the radiance of a calibrated Spectralon panel every 30 seconds throughout the period of measurements on 17th June.
5. Atmospheric measurements
The CFARR is the location of the first UK AERONET site, so measurements of aerosol optical thickness (AOT), the equivalent thickness of atmospheric water vapour (EWT) and several other atmospheric parameters were available every 15 minutes throughout the period of the experiment. In addition, three portable Microtops II sunphotometers were deployed at different locations around the test area. At one of these sites, a second Cimel sun-photometer, identical to the AERONET instrument was deployed throughout the measurement period on 17th June.
Meteorological data, including air and ground temperature, relative humidity and atmospheric pressure were recorded by the established instruments at CFARR, and the facility also provided occasional sky camera photographs and a series of water vapour profiles from an upward-looking UV lidar on 16th June. Unfortunately, this instrument was not available on 17th June.
The angular distribution of spectral irradiance (400-1000nm) was measured at frequent intervals on the 16th and 17th June using a new instrument designed by Andrew McGonigle. Click here to see a short video of Andrew describing the instrument, which is based on a temperature-stabilised miniature spectroradiometer interfaced to a telescope that can be programmed to make zenithal scans of sky irradiance. Azimuthal motion of the whole instrument was provided by manual adjustment. Each sky irradiance distribution comprised two scans: one of the whole sky, and then a second low-gain scan shortly afterwards concentrating just on the region around the Sun.
6. Biophysical data
The primary biophysical data measured were leaf area index (LAI) measurements from the various agricultural crops identified for study: wheat, barley, oats and oilseed rape. Non-destructive optical methods were used to make the LAI measurements using two Li-Cor LAI-2000 plant canopy analysers and two Delta-T SunScan canopy gap fraction measurement devices. In the wheat, barley and oats fields, five points were located in a die pattern within the field, and these were taken to define centre point of 10 m long transects, within which both the spectral and biophysical measurements were made every 1 m. The extent of the rectangular sample area within the crop was defined by the 'reach' of the various instruments, and was approximately one metre. The SunScan was much faster to use in the field, so additional measurements were made in each field by extrapolating the sampling by a further 10 m at either end of the central 10 m transect. Thus, there were approximately 50 LAI measurements from each field made using the LAI-2000, and approximately 150 LAI measurements made using the SunScan.
Leaf chlorophyll content was measured in one field (Spring barley) using a Minolta SPAD-502 instrument. Ten replicates at each of 20 locations chosen at random within each of the five transects were measured, giving a total of 100 leaf measurements.
7. Fluvial geomorphological data
 |
Researchers from the University of Southampton and the Environment Agency surveyed areas on the floodplain of the River Test as well as several cross-sections of the channel on 16th June 2006. Click here to see a short video of David Sear describing this process.
Figure 4. CASI image of one of the reaches of the River Test which were surveyed. (c) Environment Agency, 2006. |
8. On-site calibration facility
A room within the main CFARR building was converted into a calibration laboratory and equipped with the NPL Transfer Standard and Absolute Radiance Source (TSARS) (Figure 5) and a separate bench where wavelength calibrations could be performed using rare earth filters (Erbium oxide, Holmium oxide and a McCrone standard). A total of ten field spectroradiometers belonging to six groups were characterised over a period of three days by staff from the NPL and the NERC FSF. Radiance calibrations traceable to NPL standards were performed and the wavelength calibration of each instrument was checked. The experiment also involved use of a large number of reference panels, mainly composed of Spectralon, a PTFE-based material. An attempt to calibrate the panels on-site using the ASD contact probe was unsuccessful, so all the panels were sent to the NERC FSF in Edinburgh, where staff made measurements under natural solar illumination and provided the panel owners with HCRF measurements traceable to the NPL.
Figure 5. The NPL Transfer Standard Absolute Radiance Source (TSARS).
