• Korean Journal of Remote Sensing
• /
• v.34 no.5
• /
• pp.811-827
• /
• 2018

The purpose of this study is to compare machine learning algorithm and deep learning algorithm in crop classification using multi-temporal remote sensing data. For this, impacts of machine learning and deep learning algorithms on (a) hyper-parameter and (2) training sample size were compared and analyzed for Haenam-gun, Korea and Illinois State, USA. In the comparison experiment, support vector machine (SVM) was applied as machine learning algorithm and convolutional neural network (CNN) was applied as deep learning algorithm. In particular, 2D-CNN considering 2-dimensional spatial information and 3D-CNN with extended time dimension from 2D-CNN were applied as CNN. As a result of the experiment, it was found that the hyper-parameter values of CNN, considering various hyper-parameter, defined in the two study areas were similar compared with SVM. Based on this result, although it takes much time to optimize the model in CNN, it is considered that it is possible to apply transfer learning that can extend optimized CNN model to other regions. Then, in the experiment results with various training sample size, the impact of that on CNN was larger than SVM. In particular, this impact was exaggerated in Illinois State with heterogeneous spatial patterns. In addition, the lowest classification performance of 3D-CNN was presented in Illinois State, which is considered to be due to over-fitting as complexity of the model. That is, the classification performance was relatively degraded due to heterogeneous patterns and noise effect of input data, although the training accuracy of 3D-CNN model was high. This result simply that a proper classification algorithms should be selected considering spatial characteristics of study areas. Also, a large amount of training samples is necessary to guarantee higher classification performance in CNN, particularly in 3D-CNN.

• Korean Journal of Remote Sensing
• /
• v.36 no.2_1
• /
• pp.155-165
• /
• 2020

In this present study, the sensitivity of O₄ Air Mass Factor (AMF) to Aerosol Peak Height (APH) has been investigated using radiative transfer model according to various parameters(wavelength (340 nm and 477 nm), aerosol type (smoke, dust, sulfate), aerosol optical depth (AOD), surface reflectance, solar zenith angle, and viewing zenith angle). In general, it was found that O₄ AMF at 477 nm is more sensitive to APH than that at 340 nm and is stably retrieved with low spectral fitting error in Differential Optical Absorption Spectroscopy (DOAS) analysis. In high AOD condition, sensitivity of O₄ AMF on APH tends to increase. O₄ AMF at 340 nm decreased with increasing solar zenith angle. This dependency isthought to be induced by the decrease in length of the light path where O₄ absorption occurs due to the shielding effect caused by Rayleigh and Mie scattering at high solar zenith angles above 40°. At 477 nm, as the solar zenith angle increased, multiple scattering caused by Rayleigh and Mie scattering partly leads to the increase of O₄ AMF in nonlinear function. Based on synthetic radiance, APHs have been retrieved using O₄ AMF. Additionally, the effect of AOD uncertainty on APH retrieval error has been investigated. Among three aerosol types, APH retrieval for sulfate type is found to have the largest APH retrieval error due to uncertainty of AOD. In the case of dust aerosol, it was found that the influence of AOD uncertainty is negligible. It indicates that aerosol types affect APH retrieval error since absorption scattering characteristics of each aerosol type are various.

• Journal of Korean Society of Environmental Engineers
• /
• v.22 no.12
• /
• pp.2115-2123
• /
• 2000

This research was designed to investigate the mathematical model and kinetics of phenol desorption from spent activated carbon. elucidating the desorption characteristics of phenol in the case of using acetone. The Freundlich isotherm constant ($k_e$) is expressed as a function of temperature: $k_e(T)=0.1exp(797.297/T)$. The Freundlich isotherm constant(n) is a weak temperature function and is rarely affected by temperature below $50^{\circ}C$. whereas it is necessary to correct the n value with respect to temperature above $100^{\circ}C$ owing to significant deviation (~5%). Based on the assumption that the surface desorption reaction of phenol is rate limiting, the desorption model was developed. Desorption reaction constant($k_d$) was determined by means of fitting the theoretical results best to experimental ones. The Arrhenius relationships for $k_d$ was expressed by: $k_d(sec^{-1})=0.0479{\cdot}exp(-3037/T)$. The model was verified by comparing the experimental ones under different reaction conditions with the theoretical results determined by the previously estimated $k_d$. Since the difference between them is with 5%, it is expected that the desorption model of this research seems to be appropriate to explain the desorption of phenol from activated carbon by acetone. According to studies of the model. regeneration time and ratio was estimated as a function of temperature under present conditions as follows: (1) regeneration time : ${\tau}_{reg}(hr)=-0.08130T_c+8.4775$. (2) regeneration ratio : ${\eta}(%)=0.2210T_c+83.745$. The regeneration time at 15, 55, and $100^{\circ}C$. respectively. was 7, 4.2, and 0.35 hours, whereas the regeneration ratio was 87. 96. and 99%. respectively. Also. studies of the model would make it possible to determine the regeneration time and ratio under other specific conditions (temperature, applied acetone volume, amount of activated carbon, and initially adsorbed phenol amount).

• Journal of the Mineralogical Society of Korea
• /
• v.26 no.4
• /
• pp.229-244
• /
• 2013

Arsenic contamination may be brought about by a variety of natural and anthropogenic causes. Among diverse naturally-occurring chemical speciations of arsenic, trivalent (As(III), arsenite) and pentavalent (As(V), arsenate) forms have been reported to be the most predominant ones. It has been well known that the behavior of arsenic is chiefly affected by aluminum, iron, and manganese oxides. For this reason, this study was initiated to evaluate the applicability of manganese slag (Mn-slag) containing high level of Mn, Si, and Ca as an efficient sorbent of arsenic. The main properties of Mn-slag as a sorbent were investigated and the sorption of each arsenic species onto Mn-slag was characterized from the aspects of equilibrium as well as kinetics. The specific surface area and point of zero salt effect (PZSE) of Mn-slag were measured to be $4.04m^2/g$ and 7.73, respectively. The results of equilibrium experiments conducted at pH 4, 7 and 10 suggest that the sorbed amount of As(V) was relatively higher than that of As(III), indicating the higher affinity of As(V) onto Mn-slag. As a result of combined effect of pH-dependent chemical speciations of arsenic as well as charge characteristics of Mn-slag surface, the sorption maxima were observed at pH 4 for As(V) and pH 7 for As(III). The sorption of both arsenic species reached equilibrium within 3 h and fitting of the experimental results to various kinetic models shows that the pseudo-second-order and parabolic models are most appropriate to simulate the system of this study.

• Economic and Environmental Geology
• /
• v.35 no.4
• /
• pp.325-337
• /
• 2002

We examined the contamination of stream water and stream sediments by heavy metal elements with respect to distance from the abandoned Backun Au-Ag-Cu mine. High contents of heavy metals (Pb, Zn, Cu, Cd, Mn, and Fe) and aluminum in the waters connected with mining and associated deposits (dumps, tailings) reduce water quality. In the mining area, Ca and SO$_4$ are predominant cation and anion. The mining water is Ca-SO$_4$ type and is enriched in heavy metals resulted from the weathering of sulfide minerals. This mine drainage water is weakly acid or neutral (pH; 6.5-7.1) because of neutralizing effect by other alkali and alkaline earth elements. The effluent from the mine adit is also weakly acid or neutral, and contains elevated concentrations of most elements due to reactions with ore and gangue minerals in the deposit. The concentration of ions in the Backun mining water is high in the mine adit drainage water and steeply decreased award to down stream. Buffering process can be reasonably considered as a partial natural control of pollution, since the ion concentration becomes lower and the pH value becomes neutralized. In order to evaluate mobility and bioavailability of metals, sequential extraction was used for stream sediments into five operationally defined groups: exchangeable, bound to carbonates, bound to FeMn oxide, bound to organic matter, and residual. The residual fraction was the most abundant pool for Cu(2l-92%), Zn(28-89%) and Pb(23-94%). Almost sediments are low concentrated with Cd(2.7-52.8 mg/kg) than any other elements. But Cd dominate with non stable fraction (68-97%). Upper stream sediments are contaminated with Pb, and down area sediments are enriched with Zn. It is indicate high mobility of Zn and Cd.

• Journal of the Korean Magnetics Society
• /
• v.11 no.3
• /
• pp.85-96
• /
• 2001

Experimental magnetization-temperature curves for melt-spun ribbons of amorphous alloys (Dy$\_$0.33/Fe$\_$0.67/)$\_$1-x/B$\_$x/(x=0 ,0.05, 0.1, and 0.15) and (Sm$\_$0.33/Fe$\_$0.67/)$\_$1-x/B$\_$x/(x=0, 0.01, 0.02, and 0.03) (in atomic fraction) are fitted with theoretical equations based on the mean field theory in order to calculate exchange couplings between constituent elements as a function of the B content. In the case of the DyFe$_2$-B system, the sign of the exchange coupling between Dy and Fe is negative, indicating that the magnetization direction of Dy is antiparallel to that of Fe. The sign of the other two couplings are positive indicating a parallel alignment. The exchange coupling between Fe ions are greatest, while that between Dy ions is negligible. In the case of the SmFe$_2$B alloys, the sign of all the couplings are positive, indicating ferromagnetic coupling between the spins. The exchange couplings between Fe ions, and Fe and Sm are comparable to each other, but they are much greater than that between Sm ions. The high exchange coupling between Fe and Sm, which is considered to occur indirectly, is rather unexpected, but it is considered to be unique characteristics of amorphous Sm-Fe alloys. In both alloy systems, the exchange coupling between Fe ions increases with increasing B content. and this may be explained by the increase of the Fe-Fe separation with increasing B content. The exchange coupling between Fe and RE also increases with increasing B content. As the B content increases, the magnetization decreases over the whole temperature range, and the Curie temperature also decreases.

• Korean Journal of Heritage: History & Science
• /
• v.53 no.2
• /
• pp.242-253
• /
• 2020

Six ornamental metal quiver fittings were excavated from stone chamber No.1 of the ancient tombs of Jeongchon, Bokam-ri, Naju. The ornamental quiver fittings are metal, but the body of the quiver was made of organic material, so that it corroded and disappeared in the burial environment. The ornamental metal quiver fittings were made in pairs, and decorated one quiver according to the location they were found in and their forms. The ornamental metal quiver fitting can be divided into two types: A band style ornament (帶輪狀金具) which decorates the arrow pouch, and a board style ornament (板狀金具) which decorates the board connecting the waist belt. Two ornamental metal quiver fittings excavated from wooden coffin 2 of stone chamber No.1, were made in the band style, while the ornamental metal quiver fittings from southeast of stone chamber No.1 were identified as two boardstyle ornaments and two band-style ornaments for what was presumed to be belt loops. Material analysis of the ornamental metal quiver fittings shows that they are made of a gilt bronze plate attached to an iron plate, and the surface is marked with a speck of chisel to make lines and patterns. Chemical composition analysis (XRF) established that 24~40wt% Au and 50~93wt% Cu were detected on the gold surface, and it was confirmed that bronze corrosion had taken place on the gilt surface. SEM-EDS analysis of the gold plating layer identified a working line for glossing, and 7~9wt% Hg and an amalgam of gilt layers was detected, confirming the amalgam gilding. CT and FT-IR analysis established that the band style was double-layered with silk fabric under the iron plate, and there was also a lacquer piece underneath. The band-style ornaments have two layers of silk under the iron plate, along with lacquer pieces. Adding the fabric to the arrow pouch increases adhesion and decorative value. It is assumed that the lacquer pieces indicate that the surface of the lacquered arrow pouch had fallen together with the ornaments. On the other hand, the board-style ornaments have a thick layer of organic matter under the iron plate, but this is difficult to identify and appears to be a remnant of the quiver board. The characteristics of these ornamental metal quiver fittings were similar in Baekje, Silla, and Gaya cultures from the late 4th to the late 5th centuries, and enable us to identify the art of ancient gold craftwork at that time.

• Publications of The Korean Astronomical Society
• /
• v.36 no.3
• /
• pp.47-63
• /
• 2021

Variability is one of the major characteristics of Active Galactic Nuclei (AGN), and it is used for understanding the energy generation mechanism in the center of AGN and/or related physical phenomena. It it known that there exists a time lag between AGN light curves simultaneously observed at different wavelengths, which can be used as a tool to estimate the size of the area that produce the radiation. In this paper, We present long term near-infrared variability of optically bright type 1 AGN using the Wide-field Infrared Survey Explorer data. From the Milliquas catalogue v6.4, 73 type 1 QSOs/AGN and 140 quasar candidates are selected that are brighter than 18 mag in optical and located within 5 degree around the ecliptic poles. Light curves in the W1 band (3.4 ㎛) and W2 band (4.6 ㎛) during the period of 2010-2019 were constructed for these objects by extracting multi-epoch photometry data from WISE and NEOWISE all sky survey database. Variability was analyzed based on the excess variance and the probability P_var. Applying both criteria, the numbers of variable objects are 19 (i.e., 26%) for confirmed AGN and 12 (i.e., 9%) for AGN candidates. The characteristic time scale of the variability (τ) and the variability amplitude (σ) were derived by fitting the DRW model to W1 and W2 light curves. No significant correlation is found between the W1/W2 magnitude and the derived variability parameters. Based on the subsample that are identified in the X-ray source catalog, there exists little correlation between the X-ray luminosity and the variability parameters. We also found four AGN with changing W1-W2 color.