Multilayer Perceptron Neural Networks Model for Meteosat Second Generation SEVIRI Daytime Cloud Masking

Taravat, Alireza; Proud, Simon; Peronaci, Simone; Frate, Fabio Del; Oppelt, Natascha

doi:10.3390/rs70201529

Cited by 55 publications

(39 citation statements)

References 33 publications

(44 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These goals can be addressed among others, by a challenging topic of accurate daytime and night-time Aerosol Optical Depth (AOD) estimation from existing sensors on board satellites. The challenge here is not solely appropriate cloud mask filtering [26,27]. The satellite retrieval of aerosol properties over land is a difficult task, mainly due to estimation of the surface reflectance (in terms of its high temporal variability and spatial inhomogeneity) and the anisotropic bi-directional reflectance of land surfaces (inducing higher uncertainties of retrieved parameters) [28][29][30].…”

Section: Introductionmentioning

confidence: 99%

Effect of Heat Wave Conditions on Aerosol Optical Properties Derived from Satellite and Ground-Based Remote Sensing over Poland

2017

View full text Add to dashboard Cite

Abstract:During an exceptionally warm September in 2016, unique and stable weather conditions contributed to a heat wave over Poland, allowing for observations of aerosol optical properties, using a variety of ground-based and satellite remote sensors. The data set collected during 11-16 September 2016 was analysed in terms of aerosol transport (HYbrid Single-Particle Lagrangian Integrated Trajectory model (HYSPLIT)), aerosol load model simulations (Copernicus Atmosphere Monitoring Service (CAMS), Navy Aerosol Analysis and Prediction System (NAAPS), Global Environmental Multiscale-Air Quality (GEM-AQ), columnar aerosol load measured at ground level (Aerosol Robotic NETwork (AERONET), Polish Aerosol Research Network (PolandAOD)) and from satellites (Spinning Enhanced Visible and Infrared Imager (SEVIRI), Moderate Resolution Imaging Spectroradiometer (MODIS)), as well as with 24/7 PollyXT Raman Lidar observations at the European Aerosol Research Lidar Network (EARLINET) site in Warsaw. Analyses revealed a single day of a relatively clean background aerosol related to an Arctic air-mass inflow, surrounded by a few days with a well increased aerosol load of differing origin: pollution transported from Germany and biomass burning from Ukraine. Such conditions proved excellent to test developed-in-house algorithms designed for near real-time aerosol optical depth (AOD) derivation from the SEVIRI sensor. The SEVIRI AOD maps derived over the territory of Poland, with an exceptionally high resolution (every 15 min; 5.5 × 5.5 km 2 ), revealed on an hourly scale, very low aerosol variability due to heat wave conditions. Comparisons of SEVIRI with NAAPS and CAMS AOD maps show strong qualitative similarities; however, NAAPS underestimates AOD and CAMS tends to underestimate it on relatively clean days (<0.2), and overestimate it for a high aerosol load (>0.4). A slight underestimation of the SEVIRI AOD is reported for pixel-to-column comparisons with AODs of several radiometers (AERONET, PolandAOD) and Lidar (EARLINET) with high correlation coefficients (r 2 of 0.8-0.91) and low root-mean-square error (RMSE of 0.03-0.05). A heat wave driven increase of the boundary layer height of 10% is accompanied with the AOD increase of 8-12% for an urban site dominated by anthropogenic pollution. Contrary trend, with an AOD decrease of around 4% for a rural site dominated by a long-range transported biomass burning aerosol is reported. There is a positive feedback of heat wave conditions on local and transported pollution and an extenuating effect on transported biomass burning aerosol. The daytime mean SEVIRI PM2.5 converted from the SEVIRI AODs at a pixel representative for Warsaw is in agreement with the daily mean PM2.5 surface measurements, whereby SEVIRI PM2.5 and Lidar-derived Ångström exponent are anti-correlated.

show abstract

Section: Introductionmentioning

confidence: 99%

Effect of Heat Wave Conditions on Aerosol Optical Properties Derived from Satellite and Ground-Based Remote Sensing over Poland

2017

View full text Add to dashboard Cite

show abstract

“…Cerdena et al (2006Cerdena et al ( , 2007 use neural networks trained with simulated radiances for the retrieval of optical thickness, effective radius and temperature of liquid water clouds (day and night) and cirrus clouds (only day) from NOAA/AVHRR. Taravat et al (2015) use neural networks for the daytime cloud detection from SEVIRI.…”

Section: Introductionmentioning

confidence: 99%

Cirrus cloud retrieval with MSG/SEVIRI using artificial neural networks

et al. 2017

View full text Add to dashboard Cite

Abstract. Cirrus clouds play an important role in climate as they tend to warm the Earth-atmosphere system. Nevertheless their physical properties remain one of the largest sources of uncertainty in atmospheric research. To better understand the physical processes of cirrus clouds and their climate impact, enhanced satellite observations are necessary. In this paper we present a new algorithm, CiPS (Cirrus Properties from SEVIRI), that detects cirrus clouds and retrieves the corresponding cloud top height, ice optical thickness and ice water path using the SEVIRI imager aboard the geostationary Meteosat Second Generation satellites. CiPS utilises a set of artificial neural networks trained with SE-VIRI thermal observations, CALIOP backscatter products, the ECMWF surface temperature and auxiliary data.CiPS detects 71 and 95 % of all cirrus clouds with an optical thickness of 0.1 and 1.0, respectively, that are retrieved by CALIOP. Among the cirrus-free pixels, CiPS classifies 96 % correctly. With respect to CALIOP, the cloud top height retrieved by CiPS has a mean absolute percentage error of 10 % or less for cirrus clouds with a top height greater than 8 km. For the ice optical thickness, CiPS has a mean absolute percentage error of 50 % or less for cirrus clouds with an optical thickness between 0.35 and 1.8 and of 100 % or less for cirrus clouds with an optical thickness down to 0.07 with respect to the optical thickness retrieved by CALIOP. The ice water path retrieved by CiPS shows a similar performance, with mean absolute percentage errors of 100 % or less for cirrus clouds with an ice water path down to 1.7 g m −2 . Since the training reference data from CALIOP only include ice water path and optical thickness for comparably thin clouds, CiPS also retrieves an opacity flag, which tells us whether a retrieved cirrus is likely to be too thick for CiPS to accurately derive the ice water path and optical thickness.By retrieving CALIOP-like cirrus properties with the large spatial coverage and high temporal resolution of SEVIRI during both day and night, CiPS is a powerful tool for analysing the temporal evolution of cirrus clouds including their optical and physical properties. To demonstrate this, the life cycle of a thin cirrus cloud is analysed.

show abstract

“…Using more than one hidden layer (processing layer) may improve the network performance but may lead to converging to local minima. Usually, one processing layer is sufficient to classify the input data [20]. Each layer consists of one or multiple neurons depending on the solving problems.…”

Section: Classificationmentioning

confidence: 99%

Intelligent Training Algorithm for Artificial Neural Network EEG Classifications

Akkar¹,

Jasim²

2018

IJISA

View full text Add to dashboard Cite

Abstract-Artificial neural networks (ANN) have been widely used in classification. They are complicated networks due to the training algorithm used to fix their weights. To achieve better neural network performance, many evolutionary and meta-heuristic algorithms are used to optimize the network weights. The aim of this paper is to implement recently evolutionary algorithms for optimizing neural weights such as Grass Root Optimization (GRO), Artificial Bee Colony (ABC), Cuckoo Search Optimization (CSA) and Practical Swarm Optimization (PSO). This ANN was examined to classify three classes of EEG signals healthy subjects, subjects with interictal epilepsy seizure, and subjects with ictal epilepsy seizures. The above training algorithms are compared according to classification rate, training and testing mean square error, average time, and maximum iteration.

show abstract

Multilayer Perceptron Neural Networks Model for Meteosat Second Generation SEVIRI Daytime Cloud Masking

Cited by 55 publications

References 33 publications

Effect of Heat Wave Conditions on Aerosol Optical Properties Derived from Satellite and Ground-Based Remote Sensing over Poland

Effect of Heat Wave Conditions on Aerosol Optical Properties Derived from Satellite and Ground-Based Remote Sensing over Poland

Cirrus cloud retrieval with MSG/SEVIRI using artificial neural networks

Intelligent Training Algorithm for Artificial Neural Network EEG Classifications

Contact Info

Product

Resources

About