Safaa Amin

Ozone Monitoring Instrument aerosol products: a comparison study with ground-based airborne sun photometer measurements over Europe

Airborne sun photometer measurements are used to evaluate retrievals of extinction aerosol optical depth (AOD). These data are extracted from spatially coincident and temporally near-coincident measurements by the Ozone Monitoring Instrument (OMI) aboard the Aura satellite taken during 2005. OMI-measured top of atmosphere (TOA) reflectances are routinely inverted to yield aerosol products such as AOD using two different retrieval techniques: the Aura OMI Near-Ultraviolet Aerosol Data Product, OMAERUV, and the multi-wavelength Aura OMI Aerosol Data Product, OMAERO. In this work, we propose a study that specifically compares the instantaneous aerosol optical thicknesses retrieved from OMI at several locations containing sites and those of the Aerosol Robotic Network (AERONET). The result of the comparison shows that, just over Europe, OMI aerosol optical thicknesses are better retrieved in the multi-wavelength retrieval than in the near-ultraviolet. Correlations have been improved by applying a simple criterion to avoid scenes probably contaminated by thin clouds, and surface scattering. The ultraviolet irradiance positive bias in the OMI data is corrected using a procedure based on global climatological fields of aerosol absorption optical depth. The results generally show a bias significantly reduced by 5–20%, a lower variability and an unchanged, high correlation coefficient.

Ozone monitoring instrument aerosol products: Algorithm modeling and validation with ground based measurements over Europe

Airborne sun photometer measurements are used to evaluate retrievals of extinction Aerosol Optical Depth (AOD). These data are extracted from spatially coincident and temporally near coincident measurements by the Ozone Monitoring Instrument (OMI) aboard the Aura satellite during 2005. OMI measured Top Of Atmosphere (TOA) reflectances are routinely inverted to yield aerosol products such as AOD using two different retrieval techniques: a near-Ultraviolet (UV) and a multi-wavelength technique. In this work, we study the application of the two AOD modeling techniques retrieved by OMI comparing to AOD captured at several locations containing sites of the Aerosol Robotic Network (AERONET). The comparison result shows that, just over Europe, OMI aerosol optical depths are better retrieved in the multi-wavelength retrieval than in the near UV. Correlations have been improved by applying a simple criterion to avoid scenes probably contaminated by thin clouds, and surface scattering.

Enhancement of OMI aerosol optical depth data assimilation using artificial neural network

A regional chemical transport model assimilated with daily mean satellite and ground-based aerosol optical depth (AOD) observations is used to produce three-dimensional distributions of aerosols throughout Europe for the year 2005. In this paper, the AOD measurements of the Ozone Monitoring Instrument (OMI) are assimilated with Polyphemus model. In order to overcome missing satellite data, a methodology for preprocessing AOD based on neural network (NN) is proposed. The aerosol forecasts involve two-phase process assimilation and then a feedback correction process. During the assimilation phase, the total column AOD is estimated from the model aerosol fields. The main contribution is to adjust model state to improve the agreement between the simulated AOD and satellite retrievals of AOD. The results show that the assimilation of AOD observations significantly improves the forecast for total mass. The errors on aerosol chemical composition are reduced and are sometimes vanished by the assimilation procedure and NN preprocessing, which shows a big contribution to the assimilation process.

Comprehensive Survey on Dynamic Graph Models

Most of the critical real-world networks are con-tinuously changing and evolving with time. Motivated by the growing importance and widespread impact of this type of networks, the dynamic nature of these networks have gained a lot of attention. Because of their intrinsic and special characteristics, these networks are best represented by dynamic graph models. To cope with their evolving nature, the representation model must keep the historical information of the network along with its temporal time. Storing such amount of data, poses many problems from the perspective of dynamic graph data management. This survey provides an in-depth overview on dynamic graph related problems. Novel categorization and classification of the state of the art dynamic graph models are also presented in a systematic and comprehensive way. Finally, we discuss dynamic graph processing including the output representation of its algorithms.

Hybrid Randomized and Biological Preserved DNA-Based Crypt-Steganography Using Generic N-Bits Binary Coding Rule

In this paper, a blind crypto-stego technique is introduced using the cryptography and steganography concepts to achieve a double layered secured system. The proposed method consists of two phases: First is the data conversion to DNA using the proposed generic N-bits binary coding rule that leads to lower cracking probability proved by some comparative studies. Then, DNA and amino acids Playfair is applied to encrypt the DNA of the message resulting in ambiguity. In the second phase, the cipher text is placed with the ambiguity using 3:1 placement strategy. Then they are shuffled to be hidden in DNA at random positions generated by using a true real random number seed that is obtained from the atmospheric noise, thereby achieving very low cracking probability. The proposed technique is a blind preserved one as it achieves zero modification for the generated protein without extra data.

Content Based Image Retrieval with Hadoop

Hadoop has become a widely used open source framework for large scale data processing. MapReduce is the core component of Hadoop. It is this programming paradigm that allows for massive scalability across hundreds or thousands of servers in a Hadoop cluster. It allows processing of extremely large video files or image files on data nodes. This can be used for implementing Content Based Image Retrieval (CBIR) algorithms on Hadoop to compare and match query images to the previously stored terabytes of an image descriptors databases. This work presents the implementation for one of the well-known CBIR algorithms called Scale Invariant Feature Transformation (SIFT) for image features extraction and matching using Hadoop platform. It gives focus on utilizing the parallelization capabilities of Hadoop MapReduce to enhance the CBIR performance and decrease data input\output operations through leveraging Partitioners and Combiners. Additionally, image processing and computer vision tools such as Hadoop Image Processing (HIPI) and Open Computer Vision (OpenCV) are integration is shown.

Integration of Neural Network Preprocessing Model for OMI Aerosol Optical Depth Data Assimilation

A regional chemical transport model assimilated with daily mean satellite and ground based Aerosol Optical Depth (AOD) observations is used to produce three dimensional distributions of aerosols throughout Europe for the year 2005. In this paper, the AOD measurements of the Ozone Monitoring Instrument (OMI) are assimilated with Polyphemus model. In order to overcome missing satellite data, a methodology for pre-processing AOD based on Neural Network (NN) is proposed. The aerosol forecasts involve two-phase process assimilation, and then a feedback correction process. During the assimilation phase, the total column AOD is estimated from the model aerosol fields. The model state is then adjusted to improve the agreement between the simulated AOD and satellite retrievals of AOD. The results show that the assimilation of AOD observations significantly improves the forecast for total mass. The errors on aerosol chemical composition are reduced and are sometimes vanished by the assimilation procedure and NN preprocessing, which shows a big contribution to the assimilation process.

Content Based Image Retrieval Using Local Feature Descriptors on Hadoop for Indoor Navigation

This paper demonstrates Content Based Image Retrieval (CBIR) algorithms implementation on a huge image set. Such implementation will be used to match query images to previously stored geotagged image database for the purpose of vision based indoor navigation. Feature extraction and matching are demonstrated using the two famous key-point detection CBIR algorithms: Scale Invariant Feature Transformation (SIFT) and Speeded Up Robust Features (SURF). The key-points matching results using Brute Force and FLANN (Fast Library for Approximate Nearest Neighbors) on various levels for both SIFT and SURF algorithms are compared herein. The algorithms are implemented on Hadoop MapReduce framework integrated with Hadoop Image Processing Interface (HIPI) and Open Computer Vision Library (OpenCV). As a result, the experiments shown that using SIFT with KNN (4, 5, and 6) levels give the highest matching accuracy in comparison to the other methods.

Hybrid, randomized and high capacity conservative mutations DNA-based steganography for large sized data

In this paper, a well secured, high capacity, preserved algorithm is proposed through integrating the cryptography and steganography concepts with the molecular biology concepts. We achieved this by first encrypting the confidential data using the DNA Playfair cipher to avoid extra information sent to the receiver and it consequently acts as a trap for an attacker. Second, it achieves a randomized steganography process by exploiting the DNA conservative mutations. The DNA conservative mutations are utilized in a way that allows a DNA base to be substituted by another base to allow carrying two bits. Consequently, a high capacity feature is obtained with no payload for the used sequence. There are three main achieved contributions in this work. First, is hiding high capacity of data within DNA by exploiting each codon to hide two bits whilst preserving the sequence properties of protein after the steganography process, which is a trade off in the field. Secondly, using the conservative mutation with all its valid biological permutations, leads to the lowest cracking probability achieved and published till now, as proven in the security analysis section. Finally, a comparison is conducted between the proposed algorithm and five recent substitution based algorithms using large sized data up to three megabytes, to prove the algorithms scalability.

Modeling Sequence of Snapshots in Dynamic Graph

There is a significant research interest in representing evolving networks in the form of graph models. Most of the existing models use data structures to store sequence of snapshots, which are either historical or retrieved for processing purposes. Despite consuming minimal update time, these data structures induce storage redundancy, since consecutive snapshots share most of their nodes and edges in common. Compressed variants reduce this redundancy, but at the cost of increasing the update time, required to insert a new snapshot into the structure. So there is a tradeoff between storage and update time. In this paper we address the problem of storing sequence of snapshots in a compact manner while maintaining a very low update overhead. We minimize the update time by limiting the size of data that will undergo the update process. This is accomplished by separating the parts susceptible to change, from the other parts. Our experiments show that the proposed data structure achieves a compression, which is comparable to the state of the art methods, while offering an unprecedented low update overhead.