Predicting Power Conversion Efficiency of Organic Photovoltaics: Models and Data Analysis.

In this paper, the ability of three selected machine learning neural and baseline models in predicting the power conversion efficiency (PCE) of organic photovoltaics (OPVs) using molecular structure information as an input is assessed. The bidirectional long short-term memory (gFSI/BiLSTM), attentive fingerprints (attentive FP), and simple graph neural networks (simple GNN) as well as baseline support vector regression (SVR), random forests (RF), and high-dimensional model representation (HDMR) methods are trained to both the large and computational Harvard clean energy project database (CEPDB) and the much smaller experimental Harvard organic photovoltaic 15 dataset (HOPV15). It was found that the neural-based models generally performed better on the computational dataset with the attentive FP model reaching a state-of-the-art performance with the test set mean squared error of 0.071. The experimental dataset proved much harder to fit, with all of the models exhibiting a rather poor performance. Contrary to the computational dataset, the baseline models were found to perform better than the neural models. To improve the ability of machine learning models to predict PCEs for OPVs, either better computational results that correlate well with experiments or more experimental data at well-controlled conditions are likely required.

Interannual and seasonal variations in carbon exchanges over an alpine meadow in the northeastern edge of the Qinghai-Tibet Plateau, China.

The alpine meadow is highly sensitive to global climate change due to its high elevation and cold environment. To understand the dynamics of ecosystem carbon cycling, CO2 fluxes were measured over the Suli alpine meadow, which is located at the upper reach of the Shule River basin at the northeastern edge of the Qinghai-Tibet Plateau (QTP), China. The measurements were taken from October 2008 to September 2012 using the eddy covariance technique. Obvious seasonal and inter-annual variations were observed in the CO2 flux. The annual net carbon exchange ranged from -195.28 g·CO2·m-2 to -118.49 g·CO2·m-2, indicating that the alpine meadow ecosystem in this area played a role as a carbon sink. The inter-annual variability in the net carbon exchange was significantly related to the length of the growing season for the alpine meadow. The results showed that the months of June, July and August were the strongest CO2 absorption periods, while April, May and October were the strongest CO2 release periods. The annual net exchanges of CO2 in the four years were -118.49 g·CO2·m-2, -130.75 g·CO2·m-2, -195.83 g·CO2·m-2 and -160.65 g·CO2·m-2, and the average value was -151.43 g·CO2·m-2. On a seasonal scale, the monthly CO2 fluxes were largely controlled by temperature. At the annual scale, there was no dominant factor that influenced the interannual variations in the CO2 flux.

A machine learning methodology for reliability evaluation of complex chemical production systems.

System reliability evaluation is very important for safe operation and sustainable development of complex chemical production systems. This paper proposes a hybrid model for the reliability evaluation of chemical production systems. First, the influential factors in system reliability are categorized into five aspects: Man, Machine, Material, Management and Environment (4M1E), each of which represents a component subsystem of a complex chemical production process. Second, the Support Vector Machine (SVM) algorithm is used to develop machine learning models for the reliability evaluation of each subsystem, during which Particle Swarm Optimization (PSO) is applied for model parameter optimization. Third, the Random Forest (RF) algorithm is employed to correlate the reliability of the five subsystems with the reliability of the corresponding whole chemical production system. Then, Markov Chain Residual error Correction (MCRC) is adopted to improve the predictive accuracy of the machine learning model. The efficacy of the proposed hybrid model is tested on a case study, and the results indicate that the proposed model is capable of delivering satisfactory prediction results.

[Characteristics and Influencing Factors of Stable Hydrogen and Oxygen Isotopes in Groundwater in the Permafrost Region of the Source Region of the Yangtze River].

We use 84 rainfall samples collected during June to September 2017 from the Dongkemadi basin, source region of the Yangtze River, China, to analyze the characteristics and influencing factors of stable isotopes in groundwater, and further discuss the groundwater recharge sources. The results showed that the range of groundwater δ18 O values in this permafrost region varied from -15.3‰ to -12.5‰ (mean -14.0‰). The range of δD values in groundwater varied from -108.9‰ to -91.7‰ (mean -100.2‰). Compared with local atmospheric precipitation, groundwater isotopes were relatively enriched. The slope and intercept of the groundwater line (GL) in the study area were both lower than of those of the global and local meteoric water lines (GMWL and LMWL), thus indicating that groundwater in the study area was subjected to evaporation during rainfall recharge of groundwater. The d-excess values of groundwater varied from 4.9‰ to 25.0‰ (mean 11.6‰), which was close to the average d-excess value determined for global average rainfall (10‰), but lower than that of rainfall in the study area (15.1‰). The influencing factors on the composition and variation of groundwater isotopes were different in different periods. The permafrost active layer was relatively thin during periods of increasing air temperature, and groundwater isotopes were significantly affected by air temperature. A temperature decrease during the latter part of the sampling period, when the thickness of the permafrost active layer was still increasing, further increased the retention time of infiltrating rainfall in the soil, thereby eventually leading to evaporation that strengthened the enrichment of heavy isotopes in the groundwater. According to the topographic characteristics of the Dongkemadi basin, the isotopic characteristics of the groundwater, and the factors influencing the isotopic composition, we conclude that rainfall was the main source of groundwater recharge. The results of this study provide a scientific basis for studying water cycle processes in the permafrost regions of the source region of the Yangtze River.

[Characteristics of Stable Isotopes in Precipitation and Moisture Sources in the Headwaters of the Yangtze River].

Based on the stable isotopes of 73 precipitation samples continuously collected from May to October 2014 and related meteorological statistics in the Dongkemaldi Basin, the characteristics of δD, δ18O, and d-excess of precipitation, as well as the correlations between δ18O and the rainfall amount and air temperature were analyzed. The moisture sources were tracked by the HYSPLIT model to further estimate the contribution of different water vapor sources to the rainfall amount. The results showed that the range of δ18O and δD values varied from -26.5‰ to 1.9‰ and -195.2‰ to 34.0‰, respectively; meanwhile, the δ18O and δD values in precipitation fluctuated greatly with time in response to water vapor transport from different moisture sources of the Qinghai-Tibet Plateau. The slope and intercept of the Local Meteoric Water Line (LMWL) were both higher than those of the Global Meteoric Water Line (GMWL) and close to the LMWL in the northern area of the Qinghai-Tibet Plateau. The relationship between δ18O and δD in different precipitation types showed significant differences, which were mainly related to the source of water vapor and meteorological conditions during the process of precipitation formation. Because of the influence of local evaporation and the transport process of water vapor, the d-excess values of atmospheric precipitation were relatively large; the δ18O in precipitation had a significant amount effect, but had no temperature effect, thus indicating that the rainfall amount was more effective in controlling the stable isotope content of atmospheric precipitation than temperature. The modeled trajectory of vapor sources showed that water vapor of precipitation was mainly derived from the marine vapor carried by the southwest monsoon, local moisture, and the westerly water vapor, and their contributions to the rainfall amount were 43%, 36%, and 21%, respectively. The results of this study can contribute to further understanding of the atmospheric circulation characteristics and water cycle process of the Dongkemadi basin in the headwaters of the Yangtze River.

Identifying the mean residence time of soil water for different vegetation types in a water source area of the Yuanyang Terrace, southwestern China.

The mean residence time of soil water (MRTsw) for forestland and shrubland in a water source area of Yuanyang Terrace, southwestern China, was estimated using stable isotope tracer tests and the sine-wave regression model. Stable isotope analyses from precipitation and soil water were performed in 2015. The δ2H/δ18O relationship of precipitation resulted in δ2H = 7.31δ18O + 1.49, which is nearly identical to the local meteoric water line in Kunming, southwestern, China. The MRTsw was simulated at five depth ranges (0-20, 20-40, 40-60, 60-80, 80-100 cm) of the two vegetation types by precipitation δ18O input data and soil water δ18O output data. The results showed that the MRTsw values of the forestland and shrubland both increased with soil depth. However, differences in the MRTsw of the forestland (between day 53 and 94) and of the shrubland (between day 76 and 142) were discussed. Regarding the physical properties of the soil profiles from the sample plots, non-capillary porosity decreased with soil depth in the forestland (from 48.5 to 20.5 %), and was clearly higher than that in the shrubland (from 38.8 to 18.7 %). Therefore, non-capillary porosity (macropores) could be a factor that shortens the mean residence time of soil water.

[Hydrochemical Characteristics of Snow Meltwater and River Water During Snow-melting Period in the Headwaters of the Ertis River, Xinjiang].

To analyze the hydrochemical characteristics of river water and snow meltwater during snow-melting period in the Kayiertesi River, the headwaters of the Ertis River, samples of river water and meltwater were collected every day during March and April, 2014. Furthermore, the combination of descriptive statistics, Gibbs Figure and Piper Triangular diagrams of anions and cations were used for hydrochemical analyses. The results showed that the major ion compositions and hydrochemical types were significantly different between river water and snow meltwater. The total dissolved solid (TDS) in the river water ranged from 24.9 to 50.3 mg · L⁻¹. The major cations of river water were Ca²âº and Na⁺, accounting for 61% and 17% of the total cation equivalent concentration, respectively. Meanwhile, HCO3⁻ constituted about 95% of the total anions concentration. The hydrochemical type of river water was HCO3⁻-Ca²âº. The chemical composition of river water samples located in the middle with a deviation to left of Gibbs model, indicating that the major chemical process of river water was controlled by rock weath ring and precipitation but rock weathering played a more important role.

[Characteristics of Hydrogen and Oxygen Isotopes of Soil Water in the Water Source Area of Yuanyang Terrace].

Stable isotope techniques provide a new approach to study soil water movement. The precipitation and the soil water from 0 to 100 cm soil layer in 4 kinds of typical vegetation types (forest, shrub forest, grassland and non-forest land) over the water source area of Yuanyang terrace were sampled, and their isotope compositions were analyzed, aimed to understand the characteristics of stable isotopes in different depth of the soil water. The results showed that the meteoric water line in the water source area of Yuanyang terrace was δD = 6.838 4δ(18)O-5.6921 (R2 = 0.8787, n = 20), the slope and intercept were less than the global atmospheric precipitation. The hydrogen and oxygen stable isotopes in the soil water of the 4 kinds of typical types was lower than the local meteoric water line side and the fluctuation of isotope value on surface soil profile was greater. With the increasing soil depth, the fluctuation of delta 18O value was smaller and smaller, especially in the 80-100 cm soil layer which was the most obvious. The delta 18O values of the deep soil water in forest and grassland were higher than that in the surface soil. while it was on the contrary in shrub forest and non-forest land.

[Major ion chemistry of surface water in the upper reach of Shule River Basin and the possible controls ].

To analyze the major ion chemistry of water in the upper reach of the Shule River Basin and possible controls, samples of river water, groundwater, precipitation, melt water were collected and methods including descriptive statistics, Gibbs Figure, Piper Triangular diagrams of anions and cations were comprehensive used. Results showed that the major ion compositions and hydrochemical types were significantly different in different waters such as stream water, groundwater and precipitation. The total dissolved solid (TDS) in the river water ranges between 51.7 to 432. 3 mgL-1 with an average of 177.7 mgL-1. The major cations of river water are Ca2+ and Mg2+, accounting for 45% and 31% of the cations respectively. Meanwhile, HCO(3)- constituted about 75% of the anions. The hydrochemical type of river water is HCO(-)(3)-Ca2+-Mg2+. Owing to the interaction between the river and layer, the concentration of SO(2-)4 is relatively higher. Comparing major ion concentrations of the river water with local groundwater and precipitation, concentrations of the river water ranged between precipitation and groundwater but were much closer to the concentration of groundwater, indicating that the surface water was recharged by a mixture of precipitation and groundwater while groundwater is dominant. The chemical composition of surface water samples located in the middle and a bit upper of Gibbs model, which indicates that the major chemical process of river water is controlled by rock weathering and evaporation-crystallization but rock weathering plays a much more important role.

[Major ion chemistry of surface water in the Xilin River Basin and the possible controls].

Under the increasing pressure of water shortage and steppe degradation, information on the hydrological cycle in the steppe region in Inner Mongolia is urgently needed. Major ions are widely used to identify the hydrological processes in a river basin. Based on the analysis results of 239 river water samples collected in 13 sections along the Xilin River system during 2006 to 2008, combined with data from groundwater and precipitation samples collected in the same period and the meteorological and hydrological data in the Xilin River Basin, hydrochemical characteristics and the chemistry of major ions of the Xilin River water have been studied by means of Piper triangle plots and Gibbs diagrams. The results showed that: (1) the total dissolved solid (TDS) in river water mainly ranged between 136.7 mg x L(-1) and 376.5 mg x L(-1), and (2) it had an increasing trend along the river flow path. (3) The major cations and anions of river water were Ca2+ and HCO3-, respectively, and the chemical type of the river water varied from HCO3- -Ca2+ in the headwater area to HCO(3-)-Ca2+ Mg2+ in the lower part. (4) The variation in the concentration of major irons in surface water was not significant at the temporal scale. Usually, the concentration values of major irons were much higher in May than those in other months during the runoff season, while the values were a bit lower in 2007 than those in 2006 and 2008. Except for SO4(2-), the concentrations of other ions such as Ca2+, Na+, Mg2+, K+, Cl- and HCO3- showed a upward trend along the river flow path. Comparing major ion concentrations of the river water with those of local groundwater and precipitation, the concentration in river water was between those of precipitation and groundwater but was much closer to the concentration of groundwater. This indicated that the surface water was recharged by a mixture of precipitation and groundwater, and groundwater showed a larger impact. The Gibbs plot revealed that the chemical compositions of the river water were mainly affected by rock weathering in the drainage area.

[Spatial and temporal distribution of trace elements in surface water in the Xilin River Basin].

In order to better understand the hydrological process in Xilin River Basin, 248 water samples were collected in 13 sections (10 were at the mainstream and 3 were at the three tributaries) over the Xilin River during 2006-2008 and thereafter analyzed by high resolution inductively coupled plasma mass spectrometry (ICP-MS) for 20 trace elements. The temporal and spatial distribution characteristics of trace elements were obtained. The results showed that the average concentration values of trace elements were 0. 1-10 microg x L(-1). Most of those values were at the concentration ranges of precipitation and groundwater and very close to the values of groundwater, indicating that the surface water was recharged by precipitation and groundwater especially by groundwater. The variation of concentration of trace elements in surface water was not strong at the temporal scale. Usually, the concentration values of trace elements were higher in April and May than those in July and August while those values were a bit lower in 2007 than in 2006 and 2008. Most of the trace element concentrations showed a upward trend from upstream to downstream. The enrichment of trace elements was contributed to the recharge of tributaries and groundwater, the evaporation of the stream water.

[Variations and simulation of stable isotopes in precipitation in the Heihe River basin].

To study the variations of deltaD and delta18O in precipitation, 301 samples were sampled during 2002-2004 in 6 sites in the Heihe River basin, Northwestern China. The deltaD and delta18O values ranged from 59 per thousand to -254 per thousand and 6.5 per thousand to -33.4 per thousand, respectively. This wide range indicated that stable isotopes in precipitation were controlled by different condensation mechanisms as a function of air temperature and varying sources of moisture. delta18O in precipitation had a close positive relationship with the air temperature, i. e., a clear temperature effect existed in this area. At a monthly scale, no precipitation effect existed. On the other hand, a weak precipitation effect still accrued at precipitation events scale. The spatial variation of delta18O showed that the weighted average delta18O values decreased with the increasing altitude of sampling sites at a gradient of -0. 47 per thousand/100m. A regional Meteoric Water Line, deltaD = 7.82 delta18O + 7.63, was nearly identical to the Meteoric Water Line in the Northern China. The results of backward trajectory of each precipitation day at Xishui showed that the moisture of the precipitation in cold season (October to March) mainly originated from the west while the moisture source was more complicated in warm season (April to September). The simulation of seasonal delta18O variation showed that the stable isotope composition of precipitation tended to a clear sine-wave seasonal variation.

[Variations of pH value and electrical conductivity in the Dongkemadi basin, Tanggula range].

Investigation of meltwater chemistry may provide information to understand the significance of glacier in estimating of water provenance. Most notably, the role of electrical conductivity (EC) variation in meltwater during glacier melting season has attracted considerable attention, since this may reflect the water flux. Analyses for pH and EC in 229 bulk meltwater samples have provided information about water provenance at Dongkemadi Glacier basin, an outlet tongue from the Tanggula Pass, Tibetan Plateau. The samples were collected at 14:00 from 12th May to 27th September in 2005 at site of controlled meltwater flux. The results document the following findings. First phase of runoff was mainly supplied by snow and glacier ice meltwater, 31% and 65%, respectively, and the snow including fresh snow and winter deposited snow. Secondary phase of flux primarily was supplied by glacier ice meltwater and precipitation, but meltwater of frozen water in soil also has some contribute to flux, but only 2%. Third phase, ice bulk meltwater decreased and fresh snow meltwater increased. In different precipitation modalities variation trends of pH value and EC are following an orderd snow > rainwater > hailstone.