Home – Home – Jornals – Atmospheric and Oceanic Optics 2016 number 7

Evolving cognitive technologies of forest cover pattern recognition of different species and ages while hyperspectral airborne imagery processing, characteristic features of the images formation obtained by optical receiving devices are considered together with models of the registered spectra description and forest cover parameters retrieval. Specific conditions are shown of direct problems solution in the form of dependence of the spectral functional on optical properties of the forest canopy and inverse problems of the forest vegetation phytomass volume retrieval as well as its biological productivity parameters in their possible applications in climate models.

The fast atmospheric correction algorithm is described. The algorithm is based upon presentation of the reflectance for any albedo via reflectances for the same atmosphere and model surface albedos. Test results to demonstrate workability of the algorithm for high spatial resolution data processing are presented.

The inversion of the optical contrasts between the ripple and slick areas created by natural processes is analyzed. It is shown that, depending on the solar zenith angle and the angle at which spacecraft-mounted optical scanner observes the slick-ripple areas border, the intensity of light reflected from the slick area may be either greater or less than the intensity of light reflected from the ripple area. The quantitative evaluations of the sea surface slopes corresponding to inversion of the slick-ripple areas optical contrast are obtained.

The method is considered for updating of the coastal water depths according to multi- and hyperspectral remote sensing imagery in the visible and near infrared spectral range. The depth recovery for each element of the image is produced on the basis of solving the inverse problem, which consists in training an artificial neural network using a semi-analytical model of radiation transfer in water taking into account the effects of scattering and absorption of light in the underwater light field at least in three informative spectral channels for each bottom type. At the same time, it is possible to adjust the learning process by means of regression algorithms for determination of organic and mineral impurities in water on their in-situ measurements. We enriched the library of the spectral characteristics of different types of bottom and found informative signs identifying them. The results are tested according to aviation and hyperspectral space imageries.

The paper explores the features of the spectral characteristics of clean and contaminated water bodies identified on the basis of theoretical and experimental research. The original system of stable spectral features has been formed and can be used to identify and define qualitative and quantitative characteristics of sea pollution by oil products and mineral suspensions.

An indirect estimation of nitrogen and potassium on experimental agrophysical fields for the two wheat varieties "Ester" and "Trizo" by mathematical analysis of the results of aircraft hyperspectral survey in the range from 450 to 900 nm is considered. After pre-processing of the received signals and performing external calibration (atmospheric correction) of image (conversion from signal space to the spectral brightness coefficients) nitrogen and potassium content is estimated by comparison of the remote data processing results with laboratory measurements of the percentage composition of the analyzed chemical elements in plant stems on test plots. Method of subpixel analysis is used for calculations where as reference are considered two middle spectrums calculated on the plots that corresponded to the maximal content of nitrogen and potassium. The quality of the result is controlled by comparing the concentrations of these substances and estimations of concentrations for those plots that are presented in the image herewith their spectral vectors are not used as reference. The results showed a significant dependence of the accuracy of the estimations on the type of probed culture.

This paper presents a statistical microphysical model of atmospheric aerosols for the spring-summer-autumn period for the territory of Belarus and Poland developed at the base of long-term measurements at AERONET stations, carried out in Minsk and in Belsk (Poland). This model can be used for atmospheric correction of the Earth satellite monitoring data. Under this model aerosol consists of two fractions (fine and coarse), each of them having a lognormal particle size distribution, a fixed average size and rms, and fixed values of the real and imaginary parts of the refractive index. An aerosol coarse fraction consists of two components, containing spherical and non-spherical (spheroidal) particles. The only variable parameter of this model is the ratio of volume concentrations of fine and coarse fractions. It is shown that the parameters of the developed Belarus and Poland aerosol model are very close to those for the moderately absorbing aerosol model used for West Europe.

We consider an algorithm for identification of atmospheric gravity waves cloud views above water surface by MODIS imagery with 1000-m spatial resolution. Areas of the planet are outlined where these phenomena are the most frequent. Annual repeatability of atmospheric gravity waves сloud views is determined to the coasts of the Arabian Peninsula, and Australia, Mozambique Channel, and the Kurile Islands. We give a description of these phenomena identification algorithm based on Viola-Jones method. The cloud types, which forme atmospheric gravity waves cloud views are defined. The results of their identification in full-sized MODIS imagery of different planet regions are discussed.

The results of satellite differential radar interferometry in L-band (wavelength is 23 cm) are shown for the vertical land movement estimation due to seasonal deformations. A comparison was performed between results of radar interferometric measurements and field geodetic data. It was found that in winter, the discrepancy between interferometric and geodetic measurements does not exceed 1 cm, and in summer, when the moisture of topsoil changes, the discrepancy reaches 2-2.7 cm. The phase model of backscattering of microwaves by a two-layer soil with rough boundaries is proposed to assess the observed phase residuals. The model allows us to estimate the phase of backscattering wave. It is based on the ray approach and the assumption of small irregularities of the boundaries. The results of numerical calculations show that the layered structure of the earth's cover can cause a noticeable variation in the phase of backscattering wave.

The results of the remote study of small thermokarst lakes using satellite images of BKA and Alos received during summer months 2008-2014 are presented. Researches were carried out at 16 test sites in the continuous permafrost of Western Siberia. At each test site, the number of lakes, their sizes, density of lakes, limnicity of territory were determined and histograms of size distribution of lakes were built. It is shown that empirical size distribution of lakes corresponds to a power law.

Large-scale inhomogeneities of aerosol fields caused by forest fires in Siberia were studied on remote sensing data. Analysis of aerosol fields dynamics for the data on 1986, 2002, 2012 shows that formation of large-scale inhomogeneities with high AI values and lifetime of  5-7 days after forest fires cessation is possible. The spatial scales of these inhomogeneities can be several millions sq. km. Transport of emissions from forest fires in Siberia in most cases is determined by the dominant western transfer of air masses, but in some cases there was a shift in an easterly direction. Total emissions of CO₂, CO, CH₄, etc. from forest fires in Yakutia were calculated.

For two upper-air stations in Kamchatka for the period 2005-2014 the values of instability indices LIFT, TOTL and KIND, obtained by upper-air sounding and product MOD07_L2 of spectroradiometer MODIS were compared. These stations were chosen because the difference in observation time does not exceed one hour. It was found that the measurements performed at one time have a much better agreement than those different by a few hours. The most consistent and uniform were the TOTL and KIND values obtained at 12 UTC. Data from the MOD07_L2 are promising for developing new methods of diagnosis and prediction of the spatial arrangement and dynamics of convective clusters.

The paper presents a principal component (PC) method for solution of the inverse problem of retrieval of HDO to H₂O ratio in the atmosphere from spectra of outgoing Earth’s heat radiation measured with high resolution Fourier spectrometers from satellites. The method is tested by using IMG/ADEOS spectrum data over target area of Pacific Ocean. Intercomparison of results of retrieval of latitudinal distribution of HDO to H₂O ratio in the atmosphere from the IMG spectrum data using PC method with the results obtained from TES/AURA spectrum data using conventional optimal estimation methodology is presented. Good agreement between the data in Northern hemisphere and some discrepancies in Southern hemisphere are discussed.