Medium Resolution Imaging Spectrometer (MERIS) has been launched
by the European Space Agency (ESA) onboard its polar orbiting
Envisat Earth Observation Satellite.
MERIS has a high spectral and radiometric resolution and a dual spatial resolution, within a global mission covering open ocean and coastal zone waters and a regional mission covering land surfaces. One of the most outstanding features of MERIS is the programmability of its spectral bands in their width and position.
The global mission of MERIS has a major contribution to scientific projects which seek to understand the role of the oceans and ocean productivity in the climate system through observations of water colour and is furthering our ability to forecast change through models. Secondary objectives of the MERIS mission is directed to the understanding of atmospheric parameters associated with clouds, water vapour and aerosols in addition to land surface parameters, in particular vegetation processes.
With the above mentioned features MERIS meanwhile is considered to be a remote sensing tool with a large potential to contribute to climate studies and global change observations in addressing environmental features in a multi-disciplinary way.
The principal contributions of MERIS data to the study of the upper layers of the ocean are:
Apart from the above three major observable features, it should also be possible to detect special plankton blooms, for example red tides through their absorption feature near 520 nm. In addition investigations on water quality, the monitoring of extended pollution areas and topographic observations (such as coastal erosion), should also be possible.
The radiation balance of the Earth/atmosphere system is dominated by water vapour, CO2 and clouds, as well as being very dependent on the presence of aerosol. However, the global monitoring of cloud properties and their processes, is not yet sufficiently accurate. MERIS is intended to help redress this balance by providing data on cloud top height and optical thickness, water vapour column content, as well as aerosol properties.
Questions related to global change include the role of terrestrial surfaces in climate dynamics and biogeochemical cycles. Spatial and temporal models of the biosphere are currently being developed to study the mechanics of such complex systems in order to predict their behaviour under changing environmental conditions. These models are based on physical and biophysical relationships, which need to be validated on a regular basis using data from space borne sensors. Repetitive accurate physical measurements are necessary in order to quantify surface processes and to improve the understanding of vegetation seasonal dynamics and responses to environmental stress.
In order to achieve these mission goals, the different radiometric and geometric requirements imposed by the various objectives have to be satisfied. With the help of the ESA Science Advisory Group for MERIS, these requirements have been refined, taking into consideration the constraints imposed by a polar orbiting platform and the technical possibilities of an imaging spectrometer.
In advance of the launch of MERIS, the Ground Segment was designed and algorithms were developed for the interpretation of MERIS observations and dedicated studies were performed to establish the means of validating MERIS data products. This was achieved in close co-operation with the European Expert Support Laboratories whose scientists are the main authors for all retrieval algorithms. Wherever possible the underlying physical models were validated using experience acquired before Envisat launch using data provided by airborne or ship-borne campaigns and in-situ measurements on especially equipped campaign sites.
MERIS is a push broom imaging spectrometer which measures the solar reflected radiation from the Earth's in the visible and near infrared part of the spectrum during daytime.
The 1150-km-wide swath is divided into five segments covered by five identical cameras having corresponding fields of view with a slight overlap between adjacent cameras. Each camera images an across-track stripe of the Earth's surface onto the entrance slit of an imaging optical grating spectrometer. This entrance slit is imaged through the spectrometer onto a two-dimensional CCD array, thus providing spatial and spectral information simultaneously.
The spatial information along-track is determined by the push broom principle via successive read-outs of the CCD-array. Full spatial resolution data, i.e., 300 m at nadir, are transmitted over coastal zones and land surfaces. Reduced spatial resolution data, achieved by on board combination of 4 × 4 adjacent pixels across-track and along-track resulting in a resolution of approximately 1200 m at nadir, are generated continuously.
The calibration is performed at the orbital south pole, were where a reference diffuser is illuminated by the su. During calibration, the Earth-view port is closed and the sun-view port opened to provide, in the case of radiometric calibration, a uniform radiance source, and in the case of spectrometric calibration, a radiance source with a spectral signature.
In order to achieve the application objectives outlined above, the different radiometric and geometric requirements imposed by the various applications have to be satisfied. Together with the ESA Science Advisory Group for MERIS, which was created in 1988, these requirements have been refined, taking into consideration the mission constraints of a polar orbiting platform and the technical possibilities of an imaging spectrometer.
The MERIS output represents both a significant global product and data for detailed examination for regional applications. Following this requirement an operation in two spatial resolutions has been established. Full Resolution (FR) data at 300 m on-ground resolution at sub-satellite point is mainly required in coastal zones and over land. Reduced Resolution (RR) data at 1200 m on-ground resolution at sub-satellite point, is intended for large scale studies. Oceanographic and atmospheric investigations require a global Earth coverage within three days.
MERIS is designed to acquire 15 spectral bands in the 390 - 1040 NM range of the electromagnetic spectrum. The instrument has the capability to change its band position, width and gain throughout its lifetime. In accordance with the mission goals and priorities of this instrument, the following table of 15 spectral bands has been derived for oceanographic and interdisciplinary applications.
The spectral range is restricted to the visible near-infrared part of the spectrum between 390 and 1040 NM The spectral bandwidth is variable between 1.25 and 30 NM depending on the width of a spectral feature to be observed and the amount of energy needed in a band to perform an adequate observation. Over open ocean an average bandwidth of 10 NM is required for the bands located in the visible part of the spectrum. Driven by the need to resolve spectral features of the Oxygen absorption band occurring at 760 NM a minimum spectral bandwidth of 2.5 NM is required.
The spatial, spectral and radiometric programmability of MERIS is justified by the different scales of the various targets to be observed and the diversity of their spectral and radiometric properties respectively. The advantage of the programmability is not only to select width and position of a respective spectral band, but also to be able to tune the dynamic range thus make it adaptable to different target observation which may have become of a (higher) priority during the MERIS mission.
The radiometric performance is one of the most crucial requirements for MERIS because the signals coming from the ocean are weak and thus most difficult to detect and quantify. Even though the radiometrically most challenging target to be observed is the open ocean, MERIS also has to encompass a large dynamic range to cover these low level signals as well as signals emanating from bright targets such as clouds and land surfaces, throughout its spectral range. This imposes a rather demanding requirement on the MERIS radiometric performance.
In Figure 1 the different signal levels for MERIS depending on the target viewed and for the extreme orbit conditions are graphically displayed.
In the upper layer of the open ocean, the chlorophyll concentration varies from less than 0.03 mg m-3, in the oligotrophic waters, up to about 30 mg m-3 in eutrophic situations. To this variation, which spans over 3 orders of magnitude, the ocean colour responds in a non-linear way. The goal of MERIS is to discriminate 30 classes of pigment concentrations within the three orders of magnitude. The classes should be of equal logarithmic width. This requirement is translated into a radiometric sensitivity of 2 x 10-4 for NE_R (noise equivalent spectral reflectance at the sea level) set for MERIS.
For a detection of several water substances, commonly used techniques like simple colour ratios, although successfully applied for open oceans, are not sufficient. The similarity of the spectral scattering and absorption coefficients for all optically active water substances poses problems for finding an adequate procedure for their detection. Here the sun-stimulated chlorophyll fluorescence at a wavelength of 681.25 nm can improve the detection of pigment concentration. The fluorescence signal is small, but detectable from satellite. The desired spectral resolution is about 5 nm and the radiometric resolution has to be better than 0.03 Wm-2sr-1mm-1 for a discrimination of 1mg/m3 pigment concentration.
An outstanding radiometric accuracy is imperative for the atmospheric correction, which is of critical importance since typically 90% of the signal reaching the sensor originates from the atmosphere. For marine constituent detection at shorter wavelengths the atmospheric contribution of about 90 to 95% to the total signal has to be assumed for standard atmospheric conditions over the ocean (marine type aerosol and 23 km visibility), with chlorophyll concentrations of _1mg/m3. Generally a total signal of about 50 Wm-2sr-1mm-1 (including the atmosphere) has to be expected at the sensor at a solar zenith angle of 40 degrees from the sub-satellite point around 450 nm assuming low chlorophyll concentrations. These radiance levels change drastically as a function of solar elevation angles, the location of the pixel in the swath, and the wavelength position.
A knowledge of the atmospheric attenuation is required to the accuracy of about 1% or better which, for satellite (non-in situ) observations, is hardly achievable at the moment. In addition MERIS is required to have sensitivity to the polarisation of the incoming light scattered from the atmosphere lower than 1 %.
The large influence of the atmosphere on the diversity of target reflectances to be observed by MERIS necessitates a rather complex radiometric correction scheme.
1. Before a correction proper can take place the MERIS image elements, or pixels, need to be associated with the proper target algorithm processing branch, i.e. open or coastal ocean, land, atmosphere (i.e. cloud). For this the pixels are identified base on tests on spectral dependencies of the MERIS signal at specific wavelengths or an apparent barometric pressure derived from the O2 absorption.
2. The actual atmospheric correction addresses first gaseous transmittance at the same time removing the coupling between absorption and scattering, secondly the correction for the Rayleigh component and multiple scattering effects and thirdly the correction for Aerosol absorption and scattering.
3. In order to verify the corrections of the data sets various validation procedures have and are being developed using not only different radiative transfer procedures, but also different atmospheric components as input to complex simulation taking advantage of a MERIS system and data simulator. The above procedures are described in detail in contributions by Santer et al., Aiken et al., Morel et al. and Fischer et al. in IJRS MERIS Special issue Vol 20 # 9 15 June 1999.
However, atmospheric constituents are not only being treated as noise and removed by the MERIS atmospheric correction scheme, but also there is a considerable range of "by-products" leading to valuable geophysical variables such as:
- cloud top height
from the MERIS data.