Modeling hourly dissolved oxygen concentration (DO) using two different adaptive neuro-fuzzy inference systems (ANFIS): a comparative study

This article presents a comparison of two adaptive neuro-fuzzy inference systems (ANFIS)-based neuro-fuzzy models applied for modeling dissolved oxygen (DO) concentration. The two models are developed using experimental data collected from the bottom (USGS station no: 420615121533601) and top (USGS station no: 420615121533600) stations at Klamath River at site KRS12a nr Rock Quarry, Oregon, USA. The input variables used for the ANFIS models are water pH, temperature, specific conductance, and sensor depth. Two ANFIS-based neuro-fuzzy systems are presented. The two neuro-fuzzy systems are: (1) grid partition-based fuzzy inference system, named ANFIS_GRID, and (2) subtractive-clustering-based fuzzy inference system, named ANFIS_SUB. In both models, 60 % of the data set was randomly assigned to the training set, 20 % to the validation set, and 20 % to the test set. The ANFIS results are compared with multiple linear regression models. The system proposed in this paper shows a novelty approach with regard to the usage of ANFIS models for DO concentration modeling.     

Landsat and limnologically derived water quality data: A perspective

Fixed station sampling is the conventional method used to obtain data on the median water quality of reservoirs. A major source of uncertainty associated with this technique is that water quality at the fixed stations may not be representative of the ambient water quality in the reservoir at the time of sampling. This problem is particularly relevant for water quality variables such as chlorophyll, which have a markedly patchy spatial distribution. The use of Landsat reflectance data to estimate median chlorophyll concentrations in Roodeplaat Dam was investigated. A linear polynomial regression model for estimating chlorophyll concentrations from Landsat reflectance data, was firstly calibrated with chlorophyll concentration data obtained by sampling seven fixed stations on the reservoir at the time of the satellite overflight to produce an individual calibration. Secondly, the model was calibrated with a pooled set of sampled data obtained from five separate overflights, to obtain a generalised calibration.It was found that median chlorophyll concentrations determined from Landsat-derived data were similar to median chlorophyll concentrations estimated from fixed station data. However, the range of chlorophyll concentrations in the reservoir estimated from Landsat data was considerably larger than that estimated from fixed station data. Landsat derived estimates of chlorophyll concentrations have the added advantage of providing information on the spatial distribution of chlorophyll in the reservoir.     

Analysis of urban effects on soil temperature in Ankara

In this study, data from two different meteorology stations were analyzed in order to reveal the effects of the urbanization on the soil temperature. These stations are the Ankara Meteorology Station (AMS), showing the urban effects, and the Esenbo?a Meteorology Station (EMS), showing the rural effects. The soil temperatures measured at depths of 5, 10, 20, and 50 cm at 0700, 1400, and 2100 hours between 1960 and 2005 were used in the analysis. Long-term mean monthly temperatures at each depth and at each time considered were calculated and analyzed using Sen’s slope and Mann–Kendall tests. The results showed that the mean monthly urban soil temperatures were generally higher than the rural soil temperatures. The differences between temperatures measured at 5, 10, 20, and 50 cm in urban and rural stations (ΔT _s(AMS???EMS)) ranged between 1.8°C and 2.1°C. As in the urban heat islands, the differences between the urban and rural soil temperatures are high at 2100 hours and low at 1400 hours. It was also observed that, due to the increasing number of buildings around the Esenbo?a Station in recent years, the difference between the urban and rural soil temperatures seems to have become smaller. These show that the factors affecting the urban heat islands and those affecting the soil temperatures are similar. Also, the temperature differences were observed to be higher during the warm season than in the cold season. The frequency distributions of the temperature differences (ΔT _s(AMS???EMS)) reveal both positive and negative values. However, the positive temperature differences are obviously prevalent.     

Influence of anthropogenic transformations of river bed on plant and macrozoobenthos communities

This study describes the influence of urban area on plant communities and benthic invertebrates inhabiting the S?upia River (northern Poland). Ten plant communities and 37 macrozoobenthos taxa were determined during four seasonal samplings at 25 sampling sites (October 2005 and January, April, and August 2006). The obtained data set was statistically evaluated in order to reveal the influence of anthropogenic transformations on the investigated communities against the background of other abiotic factors. Multivariate regression tree (MRT) method was used for vegetation, while for benthic fauna, both MRT and artificial neural network (ANN) methods were applied. The following explanatory variables were used: season, water temperature, and salinity; location of a sampling site; degree of human impact on the riverbed; microhabitat; and substrate type. MRT analyses showed significant differences in plant community structure depending on the location of a sampling site, indicating the influence of anthropogenic pressure, while macrozoobenthos composition differed significantly only between seasons. The overall ANN model proved the importance of type and location of a sampling site for the approximation of benthic fauna density. Additionally, influence of the explanatory variables on the consecutive macrozoobenthos taxa was analyzed on the basis of separate ANN.     

Modeling of SO<Subscript>2</Subscript> distribution in Istanbul using artificial neural networks

Urban air pollution is a growing problem in developing countries. Some compounds especially sulphur dioxide (SO₂) is considered as typical indicators of the urban air quality. Air pollution modeling and prediction have great importance in preventing the occurrence of air pollution episodes and provide sufficient time to take the necessary precautions. Recently, various stochastic image-processing algorithms such as Artificial Neural Network (ANN) are applied to environmental engineering. ANN structure employs input, hidden and output layers. Due to the complexity of the problem, as the number of input–output parameters differs, ANN model settings such as the number of neurons of these layers changes. The ability of ANN models to learn, particularly capability of handling large amounts (or sets) of data simultaneously as well as their fast response time, are invariably the characteristics desired for predictive and forecasting purposes. In this paper, ANN models have been used to predict air pollutant parameter in meteorological considerations. We have especially focused on modeling of SO₂ distribution and predicting its future concentration in Istanbul, Turkey. We have obtained data sets including meteorological variables and SO₂ concentrations from Istanbul-Florya meteorological station and Istanbul-Yenibosna air pollution station. We have preferred three-layer perceptron type of ANN which consists of 10, 22 and 1 neurons for input, hidden and output layers, respectively. All considered parameters are measured as daily mean. The input parameters are: SO₂ concentration, pressure, temperature, humidity, wind direction, wind speed, strength of sunshine, sunshine, cloudy, rainfall and output parameter is the future prediction of SO₂. To evaluate the performance of ANN model, our results are compared to classical nonlinear regression methods. The over all system finds an optimum correlation between input–output variables. Here, the correlation parameter, r is 0.999 and 0.528 for training and test data. Thus in our model, the trend of SO₂ is well estimated and seasonal effects are well represented. As a result, we conclude that ANN is one of the compromising methods in estimation of environmental complex air pollution problems.     

A knowledge-based approach to regional acidification modelling

Data from over 2000 stations and knowledge from experts on atmospheric transport, soil geochemistry, lake chemistry, wetland processes and acidification modelling were assembled in an expert system. The data were grouped by aggregates of tertiary watersheds based on water chemistry knowledge. A set of expert rules was used to determine which of six existing models was most appropriate for a given set of data. Comparison of computed and observed alkalinity indicated median relative errors from 11.3–17.9%, with regression slopes ranging from 0.91–1.18 and regression coefficients between 0.82 and 0.99. The expert model performance was further confirmed with paleolimnological data and other independent sets of data. The sensitivity of the predicted alkalinity was illustrated by changing some of the rules. Given that the rules were acceptable by experts and produced reasonable agreement with observations, the knowledge-based system seemed a viable approach to the impact assessment of acidic deposition.     

Application of ANN and ANFIS models for reconstructing missing flow data

Hydrological yearbooks, especially in developing countries, are full of gaps in flow data series. Filling missing records is needed to make feasibility studies, potential assessment, and real-time decision making. In this research project, it was tried to predict the missing data of gauging stations using data from neighboring sites and a relevant architecture of artificial neural networks (ANN) as well as adaptive neuro-fuzzy inference system (ANFIS). To be able to evaluate the results produced by these new techniques, two traditionally used methods including the normal ratio method and the correlation method were also employed. According to the results, although in some cases all four methods presented acceptable predictions, the ANFIS technique presented a superior ability to predict missing flow data especially in arid land stations with variable and heterogeneous data. Comparing the results, ANN was also found as an efficient method to predict the missing data in comparison to the traditional approaches.     

Estimation of suspended sediment concentration from turbidity measurements using artificial neural networks

Suspended sediment concentration (SSC) is generally determined from the direct measurement of sediment concentration of river or from sediment transport equations. Direct measurement is very costly and cannot be conducted for all river gauge stations. Therefore, correct estimation of suspended sediment amount carried by a river is very important in terms of water pollution, channel navigability, reservoir filling, fish habitat, river aesthetics and scientific interests. This study investigates the feasibility of using turbidity as a surrogate for SSC as in situ turbidity meters are being increasingly used to generate continuous records of SSC in rivers. For this reason, regression analysis (RA) and artificial neural networks (ANNs) were employed to estimate SSC based on in situ turbidity measurements. The SSC was firstly experimentally determined for the surface water samples collected from the six monitoring stations along the main branch of the stream Harsit, Eastern Black Sea Basin, Turkey. There were 144 data for each variable obtained on a fortnightly basis during March 2009 and February 2010. In the ANN method, the used data for training, testing and validation sets are 108, 24 and 12 of total 144 data, respectively. As the results of analyses, the smallest mean absolute error (MAE) and root mean square error (RMSE) values for validation set were obtained from the ANN method with 11.40 and 17.87, respectively. However these were 19.12 and 25.09 for RA. It was concluded that turbidity could be a surrogate for SSC in the streams, and the ANNs method used for the estimation of SSC provided acceptable results.     

Detecting water quality patterns in New Hampshire’s estuaries using National Coastal Assessment probability-based survey data

Between 2000 and 2006, the New Hampshire Department of Environmental Services and the University of New Hampshire collected water quality samples at 25 to 40 stations per year in a 56.5-km² estuary as part of the Environmental Protection Agency’s National Coastal Assessment program. Due to the high density of stations, probabilistic statistics for the estuary could be calculated with low uncertainty. The proportions of the estuary exceeding thresholds in each year were calculated for temperature, salinity, dissolved oxygen, chlorophyll a, nitrogen as nitrate and nitrite, nitrogen as ammonium, phosphorus as orthophosphate, total suspended solids, and fecal coliform bacteria. These values were tested for trends over time and correlations with climate variables. The same statistical tests were applied to monthly grab sample data from a representative station in the estuary. The outcomes of the statistical tests on the two datasets were compared to determine if they provided similar information to coastal managers. Trends and correlations were equally likely to be detected using the probability-based data and the fixed station data, but the results were different for the two datasets. The differences were likely due to the distributed nature of the probability-based sampling design, which places stations in all sections of the estuary. In addition, expressing the probabilistic datasets as estimated proportions reduced variability in volatile parameters, such as bacteria, relative to the grab sample dataset. It will be important to develop tools to rectify trends from probability-based surveys with fixed station monitoring to provide clear information to managers.     

Water quality trend analysis for the Karoon River in Iran

The Karoon River basin, with a basin area of 67,000 km², is located in the southern part of Iran. Monthly measurements of the discharge and the water quality variables have been monitored at the Gatvand and Khorramshahr stations of the Karoon River on a monthly basis for the period 1967–2005 and 1969–2005 for Gatvand and Khorramshahr stations, respectively. In this paper the time series of monthly values of water quality parameters and the discharge were analyzed using statistical methods and the existence of trends and the evaluation of the best fitted models were performed. The Kolmogorov–Smirnov test was used to select the theoretical distribution which best fitted the data. Simple regression was used to examine the concentration-time relationships. The concentration-time relationships showed better correlation in Khorramshahr station than that of Gatvand station. The exponential model expresses better concentration – time relationships in Khorramshahr station, but in Gatvand station the logarithmic model is more fitted. The correlation coefficients are positive for all of the variables in Khorramshahr station also in Gatvand station all of the variables are positive except magnesium (Mg²⁺), bicarbonates () and temporary hardness which shows a decreasing relationship. The logarithmic and the exponential models describe better the concentration-time relationships for two stations.     

Modeling seasonal variations of long-term soil CO<Subscript>2</Subscript> emissions in an orchard plantation in a semiarid area,SE Turkey

Emissions of soil CO₂ under different management systems have a significant effect on the carbon balance in the atmosphere. Soil CO₂ emissions were measured from an apricot orchard at two different locations: under the crown of trees (CO₂-UC) and between tree rows (CO₂-BR). For comparison, one other measurement was performed on bare soil (CO₂-BS) located next to the orchard field. Analytical data were obtained weekly during 8 years from April 2008 to December 2016. Various environmental parameters such as air temperature, soil temperature at different depths, soil moisture, rainfall, and relative humidity were used for modeling and estimating the long-term seasonal variations in soil CO₂ emissions using two different methods: generalized linear model (GLM) and artificial neural network (ANN). Before modeling, data were randomly split into two parts, one for calibration and the second for validation, with a varying number of samples in each part. Performances of the models were compared and evaluated using means absolute of estimations (MAE), square root of mean of prediction (RMSEP), and coefficient of determination (R²) values. CO₂-UC, CO₂-BR, and CO₂-BS values ranged from 11 to 3985, from 9 to 2365, and from 8 to 1722 kg ha^?1 week^?1, respectively. Soil CO₂ emissions were significantly correlated (p?<?0.05) with some environmental variables. The results showed that GLM and ANN models provided similar accuracies in modeling and estimating soil CO₂ emissions, as the number of samples in the validation data set increased. The ANN was more advantageous than GLM models by providing a better fit between actual observations and predictions and lower RMSEP and MAE values. The results suggested that the success of environmental variables for estimations of CO₂ emissions using the two methods was moderate.     

Artificial neural network modeling of dissolved oxygen in reservoir

The water quality of reservoirs is one of the key factors in the operation and water quality management of reservoirs. Dissolved oxygen (DO) in water column is essential for microorganisms and a significant indicator of the state of aquatic ecosystems. In this study, two artificial neural network (ANN) models including back propagation neural network (BPNN) and adaptive neural-based fuzzy inference system (ANFIS) approaches and multilinear regression (MLR) model were developed to estimate the DO concentration in the Feitsui Reservoir of northern Taiwan. The input variables of the neural network are determined as water temperature, pH, conductivity, turbidity, suspended solids, total hardness, total alkalinity, and ammonium nitrogen. The performance of the ANN models and MLR model was assessed through the mean absolute error, root mean square error, and correlation coefficient computed from the measured and model-simulated DO values. The results reveal that ANN estimation performances were superior to those of MLR. Comparing to the BPNN and ANFIS models through the performance criteria, the ANFIS model is better than the BPNN model for predicting the DO values. Study results show that the neural network particularly using ANFIS model is able to predict the DO concentrations with reasonable accuracy, suggesting that the neural network is a valuable tool for reservoir management in Taiwan.     

Temporal and spatial trends in water quality of Lake Taihu,China: analysis from a north to mid-lake transect, 1991–2011

Interpretations of state and trends in lake water quality are generally based on measurements from one or more stations that are considered representative of the response of the lake ecosystem. The objective of this study is to examine how these interpretations may be influenced by station location in a large lake. We addressed this by analyzing trends in water quality variables collected monthly from eight monitoring stations along a transect from the central lake to the north in Lake Taihu (area about 2,338 km²), China, from October 1991 to December 2011. The parameters examined included chlorophyll a (Chl a), total nitrogen (TN), and total phosphorus (TP) concentrations, and Secchi disk depth (SD). The individual variables were increasingly poorly correlated among stations along the transect from the central lake to the north, particularly for Chl a and TP. The timing of peaks in individual variables was also dependent on station location, with spectral analysis revealing a peak at annual frequency for the central lake station but absence of, or much reduced signal, at this frequency for the near-shore northern station. Percentage annual change values for each of the four variables also varied with station and indicated general improvement in water quality at northern stations, particularly for TN, but little change or decline at central lake stations. Sediment resuspension and tributary nutrient loads were considered to be responsible for some of the variability among stations. Our results indicate that temporal trends in water quality may be station specific in large lakes and that calculated whole-lake trophic status trends or responses to management actions may be specific to the station(s) selected for monitoring and analysis. These results have important implications for efficient design of monitoring programs that are intended to integrate the natural spatial variability of large lakes.     

A cost-precision model for marine environmental monitoring,based on time-integrated averages

Ongoing marine monitoring programs are seldom designed to detect changes in the environment between different years, mainly due to the high number of samples required for a sufficient statistical precision. We here show that pooling over time (time integration) of seasonal measurements provides an efficient method of reducing variability, thereby improving the precision and power in detecting inter-annual differences. Such data from weekly environmental sensor profiles at 21 stations in the northern Bothnian Sea was used in a cost-precision spatio-temporal allocation model. Time-integrated averages for six different variables over 6 months from a rather heterogeneous area showed low variability between stations (coefficient of variation, CV, range of 0.6–12.4%) compared to variability between stations in a single day (CV range 2.4–88.6%), or variability over time for a single station (CV range 0.4–110.7%). Reduced sampling frequency from weekly to approximately monthly sampling did not change the results markedly, whereas lower frequency differed more from results with weekly sampling. With monthly sampling, high precision and power of estimates could therefore be achieved with a low number of stations. With input of cost factors like ship time, labor, and analyses, the model can predict the cost for a given required precision in the time-integrated average of each variable by optimizing sampling allocation. A following power analysis can provide information on minimum sample size to detect differences between years with a required power. Alternatively, the model can predict the precision of annual means for the included variables when the program has a pre-defined budget. Use of time-integrated results from sampling stations with different areal coverage and environmental heterogeneity can thus be an efficient strategy to detect environmental differences between single years, as well as a long-term temporal trend. Use of the presented allocation model will then help to minimize the cost and effort of a monitoring program.     

Probability-based nitrate contamination map of groundwater in Kinmen

Groundwater supplies over 50 % of drinking water in Kinmen. Approximately 16.8 % of groundwater samples in Kinmen exceed the drinking water quality standard (DWQS) of NO₃ ^?-N (10 mg/L). The residents drinking high nitrate-polluted groundwater pose a potential risk to health. To formulate effective water quality management plan and assure a safe drinking water in Kinmen, the detailed spatial distribution of nitrate–N in groundwater is a prerequisite. The aim of this study is to develop an efficient scheme for evaluating spatial distribution of nitrate–N in residential well water using logistic regression (LR) model. A probability-based nitrate–N contamination map in Kinmen is constructed. The LR model predicted the binary occurrence probability of groundwater nitrate–N concentrations exceeding DWQS by simple measurement variables as independent variables, including sampling season, soil type, water table depth, pH, EC, DO, and Eh. The analyzed results reveal that three statistically significant explanatory variables, soil type, pH, and EC, are selected for the forward stepwise LR analysis. The total ratio of correct classification reaches 92.7 %. The highest probability of nitrate–N contamination map presents in the central zone, indicating that groundwater in the central zone should not be used for drinking purposes. Furthermore, a handy EC–pH-probability curve of nitrate–N exceeding the threshold of DWQS was developed. This curve can be used for preliminary screening of nitrate–N contamination in Kinmen groundwater. This study recommended that the local agency should implement the best management practice strategies to control nonpoint nitrogen sources and carry out a systematic monitoring of groundwater quality in residential wells of the high nitrate–N contamination zones.     

Spatial variability of air pollutants in the city of Taranto, Italy and its potential impact on exposure assessment

Epidemiological studies typically use monitored air pollution data from a single station or as averaged data from several stations to estimate population exposure. In industrialized urban areas, this approach may present critical issues due to the spatial complexities of air pollutants which are emitted by different sources. This study focused on the city of Taranto, which is one of the most highly industrialized cities in southern Italy. Epidemiological studies have revealed several critical situations in this area, in terms of mortality excess and short-term health effects of air pollution. The aims of this paper are to study the variability of air pollutants in the city of Taranto and to interpret the results in relation to the applicability of the data in assessing population exposure. Meteorological and pollution data (SO2, NO2, PM10), measured simultaneously and continuously during the period 2006–2010 in five air quality stations, were analyzed. Relative and absolute spatial concentration variations were investigated by means of statistical indexes. Results show significant differences among stations. The highest correlation between stations was observed for PM10 concentrations, while critical values were found for NO2. The worst values were observed for the SO2 series. The high values of 90th percentile of differences between pairs of monitoring sites for the three pollutants index suggest that mean concentrations differ by large amounts from site to site. The overall analysis supports the hypothesis that various parts of the city are differently affected by the different emission sources, depending on meteorological conditions. In particular, analysis revealed that the influence of the industrial site may be primarily identified with the series of SO2 data which exhibit higher mean concentration values and positive correlations with wind intensity when the monitoring station is downwind from the industrial site. Results suggest evaluating the population exposure to air pollutants in industrialized cities by taking into account the possible zones of influence of different emission sources. More research is needed to identify an indicator, which ought to be a synthesis of several pollutants, and take into account the meteorological variables.     

Fuzzy Logic Model to Estimate Seasonal Pseudo Steady State Chlorophyll-A Concentrations in Reservoirs

A fuzzy logic model is developed to estimate pseudo steady state chlorophyll-a concentrations in a very large and deep dam reservoir, namely Keban Dam Reservoir, which is also highly spatial and temporal variable. The estimation power of the developed fuzzy logic model was tested by comparing its performance with that from the classical multiple regression model. The data include chlorophyll-a concentrations in Keban lake as a response variable, as well as several water quality variables such as PO₄ phosphorus, NO₃ nitrogen, alkalinity, suspended solids concentration, pH, water temperature, electrical conductivity, dissolved oxygen concentration and Secchi depth as independent environmental variables. Because of the complex nature of the studied water body, as well as non-significant functional relationships among the water quality variables to the chlorophyll-a concentration, an initial analysis is conducted to select the most important variables that can be used in estimating the chlorophyll-a concentrations within the studied water body. Following the outcomes from this initial analysis, the fuzzy logic model is developed to estimate the chlorophyll-a concentrations and the advantages of this new model is demonstrated in model fitting over the traditional multiple regression method.     

Potential Impact of Climate Change on Residential Energy Consumption in Dhaka City

This study aimed to assess the impacts of climate change on residential energy consumption in Dhaka city of Bangladesh. The monthly electricity consumption data for the period 2011–2014 and long-term climate variables namely monthly rainfall and temperature records (1961–2010) were used in the study. An ensemble of six global circulation models (GCMs) of coupled model intercomparison project phase 5 (CMIP5) namely, BCCCSM1-1, CanESM2, MIROC5, MIROC-ESM, MIROC-ESM-CHEM, and NorESM1-M under four representative concentration pathway (RCP) scenarios were used to project future changes in rainfall and temperature. The regression models describing the relationship between historical energy consumption and climate variables were developed to project future changes in energy consumptions. The results revealed that daily energy consumption in Dhaka city increases in the range of 6.46–11.97 and 2.37–6.25 MkWh at 95% level of confidence for every increase of temperature by 1 °C and daily average rainfall by 1 mm, respectively. This study concluded that daily total residential energy demand and peak demand in Dhaka city can increase up to 5.9–15.6 and 5.1–16.7%, respectively, by the end of this century under different climate change scenarios.     

Introduction of Relative Abundance Distribution (RAD) Indices, Estimated from the Rank-Frequency Diagrams (RFD), to Assess Changes in Community Diversity

Quantifying a relative abundance distribution based on thesampling of a set of species is a widespread problem in ecology.A number of diversity indices have been proposed and used in numerous works in spite of a lack of statistical characteristics and tests of comparison. The relative abundancedistribution can also be described using rank-frequency diagrams but fitting these diagrams to mathematical models such as the Zipf-Mandelbrot model remains problematic. Strong correlation between the Zipf-Mandelbrot model parameters prevent their estimation by optimization algorithm. In light of this, new indices of sampled communities are introduced here. These indices are two linear regression slopes estimated from rank-frequency diagrams. The numerous statistical studies that have been carried out on linear regression models are used to compare sampled communities. These new indices possesscharacteristic properties with an ecological meaning.Correlations between these indices, the Zipf-Mandelbrot modelparameters and an evenness diversity index are examined. Anecological application is made using entomological data as anexample. This example consists of a transect from the edge of apond to a dry forest along which 60,000 insects were sampledfrom six different sampling stations. Using the new indicesdescribed here, station C, located at the edge of that areasubject to influence from the pond, is differentiated from theother stations. This station would seem to present the lowestdegree of niche diversity and the lowest evenness, and recent observations confirm the deterioration of this station.     

Development of MODIS data-based algorithm for retrieving sea surface temperature in coastal waters

A new algorithm was developed for retrieving sea surface temperature (SST) in coastal waters using satellite remote sensing data from Moderate Resolution Imaging Spectroradiometer (MODIS) aboard Aqua platform. The new SST algorithm was trained using the Artificial Neural Network (ANN) method and tested using 8 years of remote sensing data from MODIS Aqua sensor and in situ sensing data from the US coastal waters in Louisiana, Texas, Florida, California, and New Jersey. The ANN algorithm could be utilized to map SST in both deep offshore and particularly shallow nearshore waters at the high spatial resolution of 1 km, greatly expanding the coverage of remote sensing-based SST data from offshore waters to nearshore waters. Applications of the ANN algorithm require only the remotely sensed reflectance values from the two MODIS Aqua thermal bands 31 and 32 as input data. Application results indicated that the ANN algorithm was able to explaining 82–90% variations in observed SST in US coastal waters. While the algorithm is generally applicable to the retrieval of SST, it works best for nearshore waters where important coastal resources are located and existing algorithms are either not applicable or do not work well, making the new ANN-based SST algorithm unique and particularly useful to coastal resource management.