Estimation of soil moisture using deep learning based on satellite data: a case study of South Korea

The Korea Meteorological Administration uses soil moisture (SM) observed by the Advanced Microwave Scanning Radiometer-2 (AMSR2) to monitor drought. However, it may not be appropriate for monitoring drought in South Korea due to significant underestimation of SM. In this study, we used a deep learning method that performs better than traditional statistical and physical models for reliable estimation of SM based on remotely sensed satellite data. For estimating SM, we carefully selected input variables that exhibit a feedback loop with SM. To build an effective deep learning model, we examined the influences of sampling criteria and input parameters as well as the accuracy of several deep neural networks. The selected model was cross-validated to determine its stability. The estimated SM using deep learning had a high correlation coefficient (R) of 0.89 and a low root mean square error (RMSE; 3.825%) and bias (?0.039%) compared to in-situ measurements. A time series analysis using dynamic time warping was conducted which showed that the estimated SM was almost similar to the in-situ SM. In order to investigate the improvement in SM estimation using our method, it was compared with the Global Land Data Assimilation System and AMSR2. Significant improvements in R and a reduction in error values by more than half were achieved using our method. The estimated SM has finer spatial resolution at 4 km, and it can be rapidly produced, which will be useful for drought monitoring over the Korean Peninsula in near-real-time.     

Evaluation of multi‐use stormwater detention basins for improved urban watershed management

Detention basins are used to capture postdevelopment runoff and control the peak discharge of the outflow using orifices and weirs. The use of detention basins is typical practice in the construction of new developments on the fringe of existing urban areas, such as the Ulsan–Hwabong district in the city of Ulsan, South Korea. In this study, the required volume and flooding area of a detention basin was determined to control development outflow peaks for 2‐year, 10‐year, and 100‐year design storms with type II rainfall distributions as characterized by the US Department of Agriculture's Soil Conservation Service method. The rainfall–runoff simulation model used was the US Environmental Protection Agency's Storm Water Management Model (EPA‐SWMM) 5, which is the latest version of the software, updated for Windows. We designed three cases of detention basins multi‐staged by 2‐year, 10‐year, and 100‐year design storms and verified the designs with the application of 49 years (1961–2009) of hourly historical rainfall data. The three detention basin designs were compared in terms of the total construction and land costs as well as the benefits associated with recreational facilities or parking lot use. As a result, the design sizes of the detention basins are slightly greater than the actual sizes needed based on the historical rainfall application. Multi‐use detention basins (MDBs) based on 2‐year and 10‐year design storms were found to yield 37.4% and 22.8% benefits, respectively, for recreational facility use compared with detention basins without multi‐use space, and the results also indicate that benefits accrue after 6.5 years for parking lot use. The results of this study suggest that an MDB based on a 2‐year design storm is the most cost‐effective design among the three cases considered for Ulsan, South Korea. Copyright © 2012 John Wiley & Sons, Ltd.     

Solubility of Fe2(OH)3Cl (pure-iron end-member of hibbingite) in NaCl and Na2SO4 brines

Pure-iron end-member hibbingite, Fe₂(OH)₃Cl(s), may be important to geological repositories in salt formations, as it may be a dominant corrosion product of steel waste canisters in an anoxic environment in Na–Cl- and Na–Mg–Cl-dominated brines. In this study, the solubility of Fe₂(OH)₃Cl(s), the pure-iron end-member of hibbingite (Fe^II, Mg)₂(OH)₃Cl(s), and Fe(OH)₂(s) in 0.04 m to 6 m NaCl brines has been determined. For the reactionFe₂(OH)₃Cl(s) + 3H⁺ ? 3 H₂O + 2 Fe²⁺ + Cl^?,the solubility constant of Fe₂(OH)₃Cl(s) at infinite dilution and 25 °C has been found to be log₁₀ K = 17.12 ± 0.15 (95% confidence interval using F statistics for 36 data points and 3 parameters). For the reactionFe(OH)₂(s) + 2H⁺ ? 2 H₂O + Fe²⁺,the solubility constant of Fe(OH)₂ at infinite dilution and 25 °C has been found to be log₁₀ K = 12.95 ± 0.13 (95 % confidence interval using F statistics for 36 data points and 3 parameters). For the combined set of solubility data for Fe₂(OH)₃Cl(s) and Fe(OH)₂(s), the Na⁺–Fe²⁺ pair Pitzer interaction parameter θ_{Na⁺/Fe²⁺} has been found to be 0.08 ± 0.03 (95% confidence interval using F statistics for 36 data points and 3 parameters). In nearly saturated NaCl brine we observed evidence for the conversion of Fe(OH)₂(s) to Fe₂(OH)₃Cl(s). Additionally, when Fe₂(OH)₃Cl(s) was added to sodium sulfate brines, the formation of green rust(II) sulfate was observed, along with the generation of hydrogen gas. The results presented here provide insight into understanding and modeling the geochemistry and performance assessment of nuclear waste repositories in salt formations.     

Mechanism of tremolite carbonation

The direct aqueous carbonation of fibrous tremolite was investigated by TEM, powder XRD, and FT-IR to elucidate the transformation mechanism of tremolite into calcite. The TEM study revealed that tremolite dissolves along the weak cleavage on (1 0 0) and (1 1 0) and the cleavage fragment fibers. The preferential release of Ca, relative to Si or Mg, into the solution resulted in the crystallization of compositionally homogeneous calcite and in the concurrent formation of proto-saponite. X-ray diffraction analysis revealed that the amount of product calcite increased rapidly within 15 min. Subsequently, after a 5 h carbonation period, it increased slowly to about 60%. The product calcite is euhedral. In contrast, the unreacted tremolite grains retain the original crystal structure even after prolonged carbonation. The morphological modification of tremolite from fibrous to round may have important implications for the elimination of its toxicity.     

Field-monitored settlement and other behavior of a multi-stage municipal waste landfill, Korea

The behavior of the Gimpo #2 landfill, which is an active landfill and the largest in Korea, is analyzed using field measurement data obtained from various field instruments installed within the landfill. The data included in this analysis are the leachate head within the landfill, waste load data using soil pressure plate and settlement data from settlement plate on the surface of the waste of each stage fill including the settlement of the soft foundation clay soil. Landfill blocks are selected both near the embankment and in the center area of the landfill. The analysis of the field-monitored data showed that the leachate head increase was negligible near the embankment. It was significant in the central block as the waste loads increase and reached 15 m at the fourth stage of waste disposal. The reason that the leachate head is higher in the central block than near the embankment is due to the long drainage path and the loss of gradient of drain pipes. The range of unit weight of the waste converted from the measurement data of earth pressure cell was 0.91–1.24 t/m³ and the average value was 1.05 t/m³. The values reflect well the waste compositions recently buried in GML #2, since from 1998 the waste disposed in GML #2 did not contain food waste. The magnitude of final settlements that occurred in each stage loading of 5 m thickness in the peripheral block was very close to 120 cm. The settlement rate of the waste by dividing the thickness of waste was 24 %. This rate can be divided into 10 % by waste loading and 14 % by waste decomposition. The delay of settlements is recognized in each waste layer for second and third loading in the central block due to the accumulation of leachate within the landfill.     

Analysis of flow behavior in a landfill with cover soil of low hydraulic conductivity

 This paper presents the results of field tests of hydrologic parameters in a landfill and the results of numerical simulation to find the efficiency of the pumping method to reduce leachate levels in the landfill. The field hydraulic conductivity and storativity of waste and buried cover soils in the landfill are measured by pumping and slug tests. The hydrologic condition inside the landfill is first calibrated using the drawdown-time curve obtained from the pumping test, and the flow behavior of leachate during pumping in the landfill, when various layers of waste and buried cover soil exist, is analyzed through three-dimensional numerical simulation of flow. The results of the field investigation show that the buried cover soil of low hydraulic conductivity forms an impermeable layer preventing the downward flow of leachate and upward flow of landfill gas. The hydraulic conductivities of the pumping test and slug tests were quite close on the same order of magnitude. It was also possible to match the drawdown-time data of the field tests with those of the model using input data close to the hydrologic property obtained from the field tests. The numerical flow analysis showed that pumping was possible up to 120 tons/day for a single well without a drain, while the pumping rate could be increased to 300 tons/day for the same well with the drain. From the vertical section of the flow vector with a horizontal drain, the barrier role of buried cover soil is identified, which was proposed by examining the water contents of the disposed cover soil and waste in the field. Received: 15 May 1998 · Accepted: 4 January 1999     

Performance of Weibull function as a diameter distribution model for <Emphasis Type="Italic">Pinus thunbergii</Emphasis> stands in the eastern coast of South Korea

This study was carried out to determine the performance of percentile-based Weibull diameter distribution model for Pinus thunbergii stands thriving along the eastern coast of South Korea. The parameter recovery technique was used to estimate the three parameters of the Weibull model. The analysis demonstrated satisfactory results based on the following test statistics for the principal percentile models: fit index (FI) range from 0.501 (minimum diameter) to 0.932 (50th diameter percentiles) and root mean square error (RMSE) range from 0.112 (quadratic mean diameter) to 3.572 (minimum diameter). The developed model was further evaluated by determining the mean bias (ē) in trees per ha (TPH) for each diameter class, and the results showed highest over-prediction in the 20 cm, and under-prediction in the 16 cm and 24 cm diameter classes. The goodness of fit tested by Kolmogorov-Smirnov (KS) test showed no significant differences (p>0.05) between the observed and predicted diameter distributions for almost all plots. Using site index and aboveground biomass (AGB) models developed for P. thunbergii in South Korea, a model to predict the AGB per ha for each diameter class and subsequently the total AGB of the stand was created. An application guide was also created, which will serve as a decision-support tool for forest managers in quantifying the future total AGB in P. thunbergii stands located in the eastern coast of South Korea and, subsequently, the quantification of potential carbon stocks aside from being a vital input in designing efficient management and protection strategies for these stands.     

通辽市R_(5d)的气候特征分析

文章利用通辽市境内科尔沁区、开鲁、鲁北、保康、甘旗卡、库伦、大沁他拉七个测站53a(1959—2011年)逐日降水量资料进行通辽市连续5d最大降水量(R5d)的气候特征分析,结果表明,通辽市R5d为减小趋势,在0.05显著性水平以上,倾向率为-4.56mm/10a,各站R5d均为减少趋势,其中库伦、开鲁、科尔沁区的倾向率为-7.0^-5.64mm/10a,减少较为明显;R/S分析结果为,全市及各站均符合赫斯特现象,H指数>0.5,未来将持续减少,其中鲁北、科尔沁区、开鲁的H指数较大,维持减少趋势将较为强劲;M-K突变检测发现,R5d在1965年发生突变,此后UF下降趋势超过了0.001极显著水平。     

初边界条件改变对四川盆地东部强降水模拟敏感性影响

利用WRF(V3.3.1)模式及其同化系统,根据不同地面观测要素同化对模式侧边界条件和初始条件引起的改变设计数值模拟方案,详细分析了WRF模式初始和侧边界条件的改变对四川盆地东部强降水模拟的敏感性影响.结果显示,WRF模式对初始和侧边界条件改变的敏感度均较明显,对地面温度变化的敏感度要高于对地面风场变化的敏感度.相对湿度场、风场和垂直速度场,均体现为高层比低层更敏感,而温度场则体现为低层比高层更敏感.在模拟时段内,侧边界条件的改变对模拟结果带来的不确定性影响随时间的增加逐渐增加,初始条件改变对模拟结果带来的不确定性影响随时间增加而逐渐减小.