Search

To improve the interpretability of estimation processes in machine learning, we applied the Shapley additive explanation method to six tree-based models for predicting downstream salinity. Decision tree-based adaptive boosting (AdaBoost) demonstrated the highest performance, followed by extreme gradient boosting (XGBoost). Global post-decomposition in XGBoost identified water level, a major cause of salinity intrusion in actual environments, as the main contributor of salinity intrusion, while an interaction effect between water temperature and sea level pressure was identified as the main contributor in AdaBoost. Local post-decomposition in XGBoost enables visual interpretation of most of the salinity behaviors. However, the interaction effect between water level and sea level pressure identified as the main contributor by XGBoost was unlikely to drive the actual increase in salinity immediately after a heavy rainfall event. We conclude that highly accurate estimations are not necessarily interpretable and reliable and local post-decomposition can detect inconsistencies in estimation processes.

To analyze physiological and biochemical marker responses of brackish water clam Corbicula japonica under global warming conditions, different water temperatures (20°C and 25°C), salinity (5 and 20 psu), and food availability (0.5 and 2.0 mg SS ind−1 d−1) levels were tested. Appropriate biomarkers were selected based on the sensitivity of biomarker responses, condition index (CI), energy reserves, superoxide dismutase (SOD), catalase, oxyradical-absorbance capacity, and lipid peroxidation (LPO) (p < 0.05), and five methods of integrated biomarker response index (IBR) were calculated. CI, SOD, and LPO were employed for IBRI, although four biochemical biomarkers were generally employed for IBR. At high water temperature (HT), all biomarkers were deleteriously affected. HT significantly affected all the IBR methods (p < 0.05), and the highest correlation was discovered in IBRI (r = 0.92, p < 0.01). The lowest and highest IBRs were observed at natural (control = 1.14) and global warming (HT = 2.43) conditions, respectively. The contribution of LPO scores was higher under HT, demonstrating poor organism health. Based on our estimates, three sensitive biomarkers could detect stress in C. japonica under global warming conditions. We conclude that the IBR method based on sensitive biomarkers should be used to quantify anthropogenic impacts on aquatic organisms, especially C. japonica.

We assessed the photosynthetic efficiency of the lab-cultured large benthic foraminifers (LBFs), Calcarina gaudichaudii and Baculogypsina sphaerulata, by means of a pulse-amplitude-modulated (PAM) fluorometer for representative photosynthetic activity; and identified an optimal light intensity and water temperature for the lab-scale cultivation. It is known that in the actual environment, the convex side of LBFs exhibit higher photosynthetic efficiency [Y(II)] than concave one. However, our experimental results from the lab setting showed that Y(II) values apparently did not differ on the basis of their having convex and concave sides. This was due to the condition between the position of the LBFs within the glass vial and light irradiation. The LBFs were exposed to eleven different light conditions for a period of 72 h. The LBFs exhibited relatively higher Y(II) at 40−120 and 240 lum/ft2, compared with that at 1150−1800 lum/ft2. Using a light intensity of 80 lum/ft2 and water temperature of 20°C to inhibit epiphyte infestation due to the growth of suspended microalgae released by the LBF hosts, LBFs were able to be cultured for 120 d under laboratory conditions, while B. sphaerulata could be cultured for 57 d under the on-site conditions in a previous study.

Field experiments were conducted at brackish-water sites (Lake Hinuma, Hinuma River, and Naka River) to investigate oxidative stress responses of a brackish-water clam, Corbicula japonica, to sediment composition and water quality. To ascertain the influence of the sediment composition, clams were subjected to three test sediments with different contents of clay/silt and organic matter from the same site. Sediment with ~33% of clay/silt content and ~10% of ignition loss exhibited decreased oxygen radical absorbance capacity (ORAC) within two weeks (p < 0.05). However, sediment with less than ~20% of clay/silt content and/or ~7% of ignition loss had no influence on ORAC (p > 0.05). To ascertain the influence of the water quality, clams were placed in baskets without sediment at four brackish-water sites. Significant differences in ORAC were observed in response to salinity and chlorophyll a (Chl a) (p < 0.05). Multiple regression analysis indicated that the levels of salinity and water temperature as well as variations in Chl a, dissolved oxygen, and turbidity for 1 d prior to the sampling date affected ORAC. We concluded that natural influencing factors need to be considered in ORAC assays for C. japonica, when anthropogenic impacts are assessed.

The design problems of electric machines are actually treated as a kind of mixed-integer problem, because the machine shapes are defined by some integer variables, such as number of slots, and the other variables, such as the tooth width, which are here called the fundamental and shape variables, respectively. To automatically solve these design problems, this article presents an automatic design method by combining the reinforcement learning and evolutionary optimization. In the proposed method, the decision process is modeled as a tree structure to seek for the fundamental variables, which are determined as a result of the tree search depending on the value functions of the nodes. Then, the shape variables are estimated from the function of the fundamental variables. These functions are constructed based on the design data, to generate which the reinforcement learning and evolutionary optimization are employed. As a result, the proposed method can automatically be adapted to unexperienced design problems through the data generation and function learning. The proposed method is applied to a design problem of a linear induction motor. It is shown that the machine designs with the prescribed performance for given specifications are automatically obtained. Moreover, it is also shown that the acceptable candidate designs can immediately be generated when the given specification is similar to the previously-solved problems by utilizing the design data generated by the past explorations.

To investigate biomarkers representing the physiological and biochemical responses of the brackish-water clam Corbicula japonica, we conducted a full factorial-design experiment to test different water-temperature levels (20 °C and 25 °C), salinity levels (5 and 20 psu), and food-availability levels (0.5 and 2.0 mg suspended solids (SS)·ind−1·d−1). Increase in water temperature significantly decreased superoxide dismutase (SOD) and catalase (CAT) activities and oxyradical-absorbance capacity (ORAC), leading to lipid peroxidation (i.e., oxidative damage). Salinity activated or inhibited these biochemical markers. Food availability supported a detoxification mechanism against oxidative stress. Principal-components and cluster analyses revealed that a total of eight experimental conditions fell into three groups related to water temperature and/or salinity. The shift from Group I (20 °C water temperature) to Group II (25 °C water temperature and 5-psu salinity) demonstrated that the condition index, SOD, CAT, and ORAC had significantly decreased. With the further shift to Group III (25 °C water temperature and 20-psu salinity), we found a prominent increase in ORAC, which led to oxidative damage but no mortality. We conclude that future habitat changes driven by global warming should be closely watched, particularly given that local anthropogenic disturbances further add to natural ones.

Abstract The purpose of this study is to predict the hydrological responses to climate change in the upstream catchment of Kase River basin having an area of 225 km2 by using MIKE SHE model and GIS. Meteorological and hydrological data for the period of 5 years were available in Kase River basin. Average daily discharge data from 5-year period (1991–1995) were used for the calibration effort, and average daily discharge was measured at Kawakami station. The Nash–Sutcliffe efficiency, NSE = 0.588, suggests that the model can predict the runoff mechanism for different periods satisfactorily. To assess the climate change impacts, two scenarios were taken under consideration. The data generated from RCM 20 (Regional Climate Model) were utilized under scenario 1 to predict future period 2081–2100 in the river basin. On the other hand, CC-scenario 2 assumes an increase in average monthly air temperature about 3 °C and no change in precipitation in future at the end of twenty-first century. The results revealed that future climate variability under scenario 1 caused an increase of 408.16 mm in surface flow; 46.77 mm in terms of base flow and interflow; and an increase of 41.74 mm actual ET (evapotranspiration). The increase in surface runoff may be mostly due to increase in precipitation. Future climate variability under scenario 2 caused a decrease of 38.43 mm in surface flow; 2.25 mm in terms of base flow and interflow; and an increase of 46.62 mm actual ET. The decrease in surface runoff may be due to unchanged precipitation and increase in temperature.