Transfer learning used to analyze the dynamic evolution of the dust aerosol

Abstract To keep the advantage of Support Vector Machine (SVM) in analyzing the dynamic evolution of the dust aerosol, we introduce transfer learning as a new method because transfer learning can utilize knowledge from previously collected data and add dozens of new samples, which can significantly improve dust and cloud classification results. It can also reduce the time of sample collection and make learning efficient. In this paper, we receive significant improvement effect using SVM as the basic learner in TrAdaBoost during four consecutive dust storm days, and correct one error classification in PDF. As a result, dust aerosol in high altitude can even spread to stratosphere. Moreover, in the process of dust aerosol transportation, it is highly affected by anthropogenic aerosol, for example, the color ratio (CR) changes from 0.728 to 0.460 and finally reaches 0.466, while depolarization ratio (DR) changes from 0.308 to 0.081 and finally reaches 0.156. It is indicated that the big size and non-spherical aerosol particles reduce obviously after dust aerosol deposition, but small size and spherical anthropogenic aerosol also produce a certain effect, and on March 22, 2010 had a small recovery above the ocean following the reduction of DR and CR. Due to the MODIS resolution not meeting the observation requirement and layer identification being different between CALIPSO and CloudSat, a problem such as stratocumulus cloud in low altitude still exists in aerosol and cloud classification. Lack of ground-based auxiliary data is the main problem which hinders our validation and quantitative analysis. It is pressing for a solution in future.

[1]  Dong L. Wu,et al.  Transpacific transport and evolution of the optical properties of Asian dust , 2013 .

[2]  David M. Winker,et al.  Fully automated analysis of space-based lidar data: an overview of the CALIPSO retrieval algorithms and data products , 2004, SPIE Remote Sensing.

[3]  Patrick Minnis,et al.  Possible influences of Asian dust aerosols on cloud properties and radiative forcing observed from MODIS and CERES , 2006 .

[4]  Jimin Sun,et al.  Spatial and temporal characteristics of dust storms in China and its surrounding regions, 1960-1999 : Relations to source area and climate , 2001 .

[5]  Jin-Seok Han,et al.  Characteristics of Asian aerosol transport simulated with a regional‐scale chemical transport model during the ACE‐Asia observation , 2004 .

[6]  David M. Winker,et al.  The CALIPSO mission , 2003, IGARSS 2003. 2003 IEEE International Geoscience and Remote Sensing Symposium. Proceedings (IEEE Cat. No.03CH37477).

[7]  Liangpei Zhang,et al.  A Support Vector Conditional Random Fields Classifier With a Mahalanobis Distance Boundary Constraint for High Spatial Resolution Remote Sensing Imagery , 2014, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[8]  François Dulac,et al.  An additional low layer transport of Sahelian and Saharan dust over the north-eastern Tropical Atlantic , 1995 .

[9]  Jianping Huang,et al.  Discrimination and validation of clouds and dust aerosol layers over the Sahara desert with combined CALIOP and IIR measurements , 2014, Journal of Meteorological Research.

[10]  Ikuko Mori,et al.  Change in size distribution and chemical composition of kosa (Asian dust) aerosol during long-range transport , 2003 .

[11]  Ina Tegen,et al.  Modeling the mineral dust aerosol cycle in the climate system , 2003 .

[12]  David M. Winker,et al.  Use of probability distribution functions for discriminating between cloud and aerosol in lidar backscatter data , 2004 .

[13]  Wei Gong,et al.  New dust aerosol identification method for spaceborne lidar measurements , 2011 .

[14]  Lorenzo Bruzzone,et al.  Classification of hyperspectral remote sensing images with support vector machines , 2004, IEEE Transactions on Geoscience and Remote Sensing.

[15]  Fuzhong Weng,et al.  Detection of Asia dust storms using multisensor satellite measurements , 2007 .

[16]  Jie Xuan,et al.  Characterization of sources and emission rates of mineral dust in Northern China , 2002 .

[17]  Wei Gong,et al.  Simple multiscale algorithm for layer detection with lidar. , 2011, Applied optics.

[18]  Liangpei Zhang,et al.  Multiagent Object-Based Classifier for High Spatial Resolution Imagery , 2014, IEEE Transactions on Geoscience and Remote Sensing.

[19]  R. Arimoto,et al.  Concentration, size-distribution and deposition of mineral aerosol over Chinese desert regions , 1998 .

[20]  Chen Weinan,et al.  DUST FALL IN THE TAKLA MAKAN DESERT OF CHINA , 1999 .

[21]  Patrick Minnis,et al.  Satellite‐based assessment of possible dust aerosols semi‐direct effect on cloud water path over East Asia , 2006 .

[22]  David M. Winker,et al.  The CALIPSO Lidar Cloud and Aerosol Discrimination: Version 2 Algorithm and Initial Assessment of Performance , 2009 .

[23]  Xiaoping Liu,et al.  Knowledge transfer and adaptation for land-use simulation with a logistic cellular automaton , 2013, Int. J. Geogr. Inf. Sci..

[24]  Yingying Ma,et al.  Evaluating the Performance of SVM in Dust Aerosol Discrimination and Testing its Ability in an Extended Area , 2012, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[25]  Q. Fu,et al.  For the depolarization of linearly polarized light by smoke particles , 2013 .

[26]  Jean-Pierre Blanchet,et al.  A simulated climatology of Asian dust aerosol and its trans-Pacific transport. Part I: Mean climate and validation , 2006 .

[27]  Yoav Freund,et al.  Experiments with a New Boosting Algorithm , 1996, ICML.

[28]  Liangpei Zhang,et al.  Cloud amount and aerosol characteristic research in the atmosphere over Hubei province, China , 2009, 2009 IEEE International Geoscience and Remote Sensing Symposium.

[29]  Johannes R. Sveinsson,et al.  Spectral and spatial classification of hyperspectral data using SVMs and morphological profiles , 2008, 2007 IEEE International Geoscience and Remote Sensing Symposium.

[30]  Olga V. Kalashnikova,et al.  CALIPSO Observations of Transatlantic Dust: Vertical Stratification and Effect of Clouds , 2012 .

[31]  D. Winker,et al.  The CALIPSO Automated Aerosol Classification and Lidar Ratio Selection Algorithm , 2009 .

[32]  Pinhas Alpert,et al.  Vertical distribution of Saharan dust based on 2.5-year model predictions , 2004 .

[33]  Richard Arimoto,et al.  Dust emission from Chinese desert sources linked to variations in atmospheric circulation , 1997 .

[34]  Paul M. Mather,et al.  Support vector machines for classification in remote sensing , 2005 .

[35]  David M. Winker,et al.  Fully Automated Detection of Cloud and Aerosol Layers in the CALIPSO Lidar Measurements , 2009 .

[36]  Patrick Minnis,et al.  Long-range transport and vertical structure of Asian dust from CALIPSO and surface measurements during PACDEX , 2008 .

[37]  Zhaoyan Liu,et al.  3D structure of Asian dust transport revealed by CALIPSO lidar and a 4DVAR dust model , 2008 .

[38]  Hussein T. Mouftah,et al.  A review of techniques and technologies for sand and dust storm detection , 2012, Reviews in Environmental Science and Bio/Technology.

[39]  David M. Winker,et al.  The CALIPSO mission: spaceborne lidar for observation of aerosols and clouds , 2003, SPIE Asia-Pacific Remote Sensing.

[40]  B. Holben,et al.  Validation of MODIS aerosol optical depth retrieval over land , 2002 .

[41]  Qiang Yang,et al.  Boosting for transfer learning , 2007, ICML '07.

[42]  Zhaoyan Liu,et al.  Asian dust transported one full circuit around the globe , 2009 .

[43]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[44]  Yongxiang Hu,et al.  Occurrence, liquid water content, and fraction of supercooled water clouds from combined CALIOP/IIR/MODIS measurements , 2010 .

[45]  Patrick Minnis,et al.  Dust aerosol effect on semi-arid climate over Northwest China detected from A-Train satellite measurements , 2010 .

[46]  Dongfang Wang,et al.  Characterization of soil dust aerosol in China and its transport and distribution during 2001 ACE‐Asia: 1. Network observations , 2003 .

[47]  Zhaoyan Liu,et al.  Airborne dust distributions over the Tibetan Plateau and surrounding areas derived from the first year of CALIPSO lidar observations , 2008 .

[48]  David M. Winker,et al.  Summer dust aerosols detected from CALIPSO over the Tibetan Plateau , 2007 .

[49]  Simone Tanelli,et al.  CloudSat mission: Performance and early science after the first year of operation , 2008 .

[50]  Alexander Smirnov,et al.  Analysis of measurements of Saharan dust by airborne and ground-based remote sensing methods during the Puerto Rico Dust Experiment (PRIDE) , 2003 .

[51]  David M. Winker,et al.  CALIPSO lidar observations of the optical properties of Saharan dust: A case study of long‐range transport , 2008 .

[52]  Patrick Minnis,et al.  Detection of dust aerosol by combining CALIPSO active lidar and passive IIR measurements , 2010 .

[53]  Patrick Minnis,et al.  Taklimakan dust aerosol radiative heating derived from CALIPSO observations using the Fu-Liou radiation model with CERES constraints , 2009 .

[54]  K. V. S. Badarinath,et al.  Long-range transport of dust aerosols over the Arabian Sea and Indian region — A case study using satellite data and ground-based measurements , 2010 .

[55]  Qiang Yang,et al.  A Survey on Transfer Learning , 2010, IEEE Transactions on Knowledge and Data Engineering.

[56]  Simone Mantovani,et al.  Aerosol Optical Thickness Retrieval from Satellite Observation Using Support Vector Regression , 2010, CIARP.

[57]  Liangpei Zhang,et al.  A Hybrid Object-Oriented Conditional Random Field Classification Framework for High Spatial Resolution Remote Sensing Imagery , 2014, IEEE Transactions on Geoscience and Remote Sensing.