EarthCARE Preparation

The ESA’s cloud and aerosol mission EarthCARE will provide measurements from active sounder and passive imager from one platform. The active backscatter lidar (ATLID) will provide vertical profiles of cloud and aerosol parameters with high spatial resolution. The lidar instrument measures in nadir view, while the passive multi-spectral imager (MSI) has a swath of 150km and a pixel size of 500m. State of the art algorithms have been developed to retrieve MSI cloud products, e.g. cloud mask, ISCCP cloud types, cloud phase, cloud optical thickness, cloud effective radius and cloud top height. Based on the stand alone retrievals we combine active and passive measurements to derive a synergetic cloud product, e.g. the cloud top height. The cloud top retrieval from MSI will be an infrared effective radiating height, which is located somewhere within the cloud. The ATLID cloud top height will give the physical boundaries of the clouds alongtrack which will be used to study the relationships between the effective and true cloud top height. The synergetic cloud top height product relies on the assumption that clouds from the same ISCCP type often share geometric and microphysical properties. Therefore an cluster search algorithm has been developed to identify connected cloudfields. For such a cluster a simplified ATLID cloud profile is used for the forward model (RTTOV) to retrieve an improved cloud top height at the swath of the MSI. To verify the method we used MODIS and Calipso observations as a test bed. From the CALIOP lidar onboard CALIPSO the 1064-nm elastic backscatter signal is valid as a substitute for the ATLID 355-nm co-polar Mie signal. The 1064-nm signal has a negligible Rayleigh contribution and thus a similar shape in the presence of clouds as the Rayleighfree filtered ATLID 355-nm co-polar Mie signal.

The European Space Agency (ESA) is presently implementing the Earth Cloud, Aerosol and Radiation Explorer (EarthCARE) mission in cooperation with the Japan Aerospace Exploration Agency (JAXA). The satellite payload consists of two active and two passive instruments. The Atmospheric Lidar (ATLID) operates at 355nm and is equipped with a high-spectral resolution receiver and depolarisation channel that separates molecular from particulate back-scatter and distinguishes cloud and aerosol types. The Japanese Cloud Profiling Radar (CPR) is a highly sensitive W-band Doppler radar (94GHz) that measures cloud profiles, precipitation and vertical motion within clouds. Due to its significantly higher sensitivity when compared to CloudSat, it will detect substantially thinner ice clouds and stratocumulus. The Doppler observation will observe vertical motion in clouds providing novel information on convection, precipitating ice particles and raindrop fall speed.

A Multi-Spectral Imager (MSI) with a 150km wide swath and seven channels in the visible, near-IR, short-wave IR, and thermal IR, will provide scene context information and allow the reconstruction of three-dimensional atmospheric scenes when combined with lidar and radar retrievals. A Broad-Band Radiometer (BBR) observing broad-band solar and thermal radiation reflected and emitted from the Earth, with three fixed field of view looking forward, nadir and backward, will make collocated measurements of the outgoing reflected solar and emitted thermal radiation.

The synergistic exploitation of the four instruments will provide 3D cloud-aerosol-precipitation scenes, with collocated broad-band radiation data, over a mission lifetime of three years. The satellite acceptance review is scheduled for 2018.

This presentation will provide an overview of the mission and its expected science data products.

The EarthCARE (Earth Clouds and Radiation Explorer) mission, to be launched in 2018, will be equipped with an active atmospheric lidar (ATLID) and a cloud profiling radar (CPR), as well as a passive multi-spectral imager (MSI) and a broad-band radiometer (BBR).The overarching goal is to predict top-of-atmosphere (TOA) short-wave (SW) fluxes, based on 1D or 3D radiative transfer simulations acting on retrieved vertical profiles of cloud and atmospheric parameters (inferred from MSI, ATLID and CPR), with an accuracy of at least 10 W/m² over a footprint of (10km)², when compared against fluxes, estimated from BBR-based TOA SW radiance measurements.

Estimation of latter SW flux from SW radiances is problematic, as various atmospheric and surface conditions can lead to very different bi-directional reflective (BR) distributions of SW radiation. Clear-sky conditions form the extreme of this problematic nature. To estimate TOA SW clear-sky fluxes successfully, knowledge about the type and state of the surface as well as atmospheric constituents is necessary. Therefore, it is foreseen to employ ECWMF (European Centre for Medium-Range Weather Forecasts) IFS (Integrated Forecasting System) data in the EarthCARE mission. In addition, aerosol and surface BR funcion (BRF) climatologies will serve as auxiliary data to estimation. In contrast to support estimation by MSI data, a stand-alone TOA SW flux product is desirable.

In order to ensure reliability of EarthCARE's SW radiance-to-flux conversion under clear-sky conditions , we extracted CERES (Clouds and the Earth's Radiant Energy System) observed TOA SW radiances and estimated TOA SW fluxes for a BBR-like viewing geometry, several geophysical parameters from ECMWF ERA20C reanalysis, as well as AeroCom and MODIS (Moderate-resolution Imaging Spectroradiometer) based BRF climatologies.

We tested both linear and non-linear models to estimate TOA SW fluxes. As multi-variate linear models lack ability to select parameters, we used Genetic Algorithms to determine surface-specific optimal parameter sets. We predicted clear-sky TOA SW fluxes globally with a root-mean square error (RMSE) of 7.60 W/m². In addition, we presented insights on parameter importance.

Random Forests, i.e. multiple decision trees, presents a non-linear method which includes parameter selection. Using a Random Forest with 20 trees, we outperformed linear models by about 30% (RMSE of 5.00 W/m²). We compared parameters, selected for decision trees, with those chosen by Genetic Algorithms. Finally, we utilized knowledge on parameter importance to train Artificial Neural Networks, as a second non-linear model, for TOA SW flux estimation. We showed how prediction performance changed when using all available parameters versus a meaningful parameter subset.

 ATLID stands for “ATmospheric LIDar” and is the lidar to be flown on the Earth Clouds and Radiation Explorer (EarthCARE) platform.  ATLID is a High-Spectral Resolution (HSRL) system operating at 355nm which will enable it to to estimate both cloud/aerosol-extinction and backscatter. However, two issues that must be addressed in order to deliver reliable estimates of cloud/aerosol extinction and backscatter at the 1-km horizontal scale are low SNR (compared to terrestial lidars) and the occurrence of multiple-scattering (MS). As part of the EarthCARE algorithm development activities a novel approach for the retrieval of extinction and backscatter which addresses these two issues has been formulated. In essence, a hybrid approach which combines the favorable characteristics of elastic lidar type inversion methods with information derived from the HSRL Rayleigh channel while efficiently  accounting for the effects of  MS has been formulated. In this presentation, the algorithm will be described both within the wider context of the EarthCARE processing chain and in specific detail. Specific attention will be focused on the treatment of multiple-scattering. The operation of the algorithm will be illustrated using examples generated by the application of the Earth Care Simulator (ECSIM).

The Earth Clouds, Aerosol and Radiation Explorer (EarthCARE) is a combined ESA/JAXA mission to be launched in 2018.  EarthCARE will study the spatial (3D) distribution of clouds and aerosols and their impact on the Earth's radiative balance. To do this, the EarthCARE platform will carry a combination of active and passive sensors, with one of them the High Spectral Resolution Lidar (HSRL) ATLID operating at 355nm. A HSRL system can separate the backscatter signals from particles (Mie channel) and molecular return (Rayleigh channel) enabling the retrieval of extinction and backscatter directly.

In order to be able to derive reliable extinction and backscatter profiles, as well as a target classification (ice cloud, liquid cloud or aerosol layer etc.); an accurate feature mask is essential. A feature mask identifies 'significant returns' in the lidar signal  indicating  the presence of either aerosols or cloud particles within the beam. It does not specify the nature of the feature yet but enables the use of variable masks and binning strategies, e.g. use only those profiles which are sure to have no clouds to derive the mean aerosol signals, and calculation of feature fractions which will result in a better determination of higher order products.

In this work, the Feature Mask algorithm for the EarthCARE high spectral resolution lidar, being developed within the ESA sponsored APRIL study, is discussed.

The algorithm defines the feature mask based on the correlation of the data without relying on a number of hard coded or input dependent thresholds. As the signal strength of aerosol or very optically thin ice clouds on the single shot grid can be comparable to the instrument noise levels it was chosen to rely on image reconstruction techniques and not on signal to noise ratios and thresholds. The main reason why an image reconstruction technique is so effective for the EarthCARE lidar data is based on the principle that the signals in the Mie channel contain only particle backscatter, background noise and noise due to the Mie-Rayleigh cross-talk and no molecular backscatter.  It also ensures the derivation of a feature mask on the single shot resolution instead of directly going to a lower horizontal resolution of 1 or 10 km.

The algorithm and results for a number of different scenes including ice clouds, liquid clouds and aerosols will be presented.