Modelling geosmin concentrations in three sources of raw water in Quebec, Canada

The presence of off-flavour compounds such as geosmin, often found in raw water, significantly reduces the organoleptic quality of distributed water and diverts the consumer from its use. To adapt water treatment processes to eliminate these compounds, it is necessary to be able to identify them quickly. Routine analysis could be considered a solution, but it is expensive and delays associated with obtaining the results of analysis are often important, thereby constituting a serious disadvantage. The development of decision-making tools such as predictive models seems to be an economic and feasible solution to counterbalance the limitations of analytical methods. Among these tools, multi-linear regression and principal component regression are easy to implement. However, due to certain disadvantages inherent in these methods (multicollinearity or non-linearity of the processes), the use of emergent models involving artificial neurons networks such as multi-layer perceptron could prove to be an interesting alternative. In a previous paper (Parinet et al., Water Res 44: 5847-5856, 2010), the possible parameters that affect the variability of taste and odour compounds were investigated using principal component analysis. In the present study, we expand the research by comparing the performance of three tools using different modelling scenarios (multi-linear regression, principal component regression and multi-layer perceptron) to model geosmin in drinking water sources using 38 microbiological and physicochemical parameters. Three very different sources of water, in terms of quality, were selected for the study. These sources supply drinking water to the Québec City area (Canada) and its vicinity, and were monitored three times per month over a 1-year period. Seven different modelling methods were tested for predicting geosmin in these sources. The comparison of the seven different models showed that simple models based on multi-linear regression provide sufficient predictive capacity with performance levels comparable to those obtained with artificial neural networks. The multi-linear regression model (R ²?=?0.657, <0.001) used only four variables (phaeophytin, sum of green algae, chlorophyll-a and potential Redox) in comparison with ten variables (potassium, heterotrophic bacteria, organic nitrogen, total nitrogen, phaeophytin, total organic carbon, sum of green algae, potential Redox, UV absorbance at 254 nm and atypical bacteria) for the best model obtained with artificial neural networks (R ²?=?0.843).     

苏州市太湖饮用水源地异味物质种类及其与环境因子相关性分析

于2013年每月测定了苏州太湖饮用水源地9种异味物质含量及TP、TN、水温、藻密度等指标,结果表明,苏州太湖饮用水源地主要检出的异味物质有2-甲基异崁醇、土臭素、β-环柠檬醛和β-紫罗兰酮等4种。2-甲基异崁醇年均值是其嗅觉阈值的2.9倍,水体主要表现为土霉味。根据异味物质与环境因子相关性分析结果及相关文献讨论,水温是影响异味物质含量的重要因素。微囊藻等藻类对β-环柠檬醛和β-紫罗兰酮有明显贡献。     

固相微萃取-气相色谱-质谱法测定饮用水中土嗅素和2-甲基异茨醇

采用固相微萃取法富集,气相色谱-质谱联用法定性和定量测定饮用水中致嗅物质土嗅素和2-甲基异茨醇。研究并讨论优化了纤维头的类型、盐的种类和浓度、温度、萃取时间等因素对异味化合物萃取量的影响。土嗅素和2-甲基异茨醇的检出限分别为1.02、2.13 ng/L,相对标准偏差分别为4.96%、7.74%。2种异味化合物在5~1 000 ng/L的范围内线性关系良好,相关系数均大于0.985。因此,用该方法能够很好地分析水中痕量的异味化合物。     

Seasonal change and correlation with environmental parameters for 2-MIB in Feng-Shen Reservoir, Taiwan

Two common drinking water earthy/musty odorants, 2-methyl-isoborneol (2-MIB) and trans-1,10-dimethyl-trans-9-decalol (geosmin), were monitored and analyzed for their correlation with environmental conditions in Feng-Shen Reservoir (FSR), Taiwan. Long-term monitoring results, from December 2000 to July 2003, indicated that 2-MIB was present in the reservoir at concentrations between 10 and 200 ng L(-1), and those of geosmin were always smaller than 10 ng L(-1). The 2-MIB concentrations followed a trend of higher concentrations during the warm seasons. After being analyzed with 10 water quality and three meteorological parameters, 2-MIB concentration was found to correlate with corresponding air and water temperatures. An analysis of air temperature history versus 2-MIB concentration in FSR suggested that 2-MIB concentration was best correlated with the mean of average daily temperature from 1 to 9 days before sampling, with R (2) = 0.90. This correlation was further employed to predict the 2-MIB concentration observed during another sampling period in FSR. A similar degree of fit was observed between predictions and experimental data, suggesting the potential applicability of the approach. In addition to the effect of temperature, heavy rainfalls that occurred before the sampling time may also reduce the 2-MIB concentration in the reservoir. A short-term continuous monitoring of the odorants for 32 h showed that both geosmin and 2-MIB concentrations in the reservoir remained almost constant, with only about a 10% difference during the sampling period. This may suggest that the odorants were uniformly distributed in the water near the sampling location in the reservoir.     

Water quality time series for Big Melen stream (Turkey): its decomposition analysis and comparison to upstream

Big Melen stream is one of the major water resources providing 268 km³ year^???1 of drinking and municipal water for Istanbul. Monthly time series data between 1991 and 2004 for 25 chemical, biological, and physical water properties of Big Melen stream were separated into linear trend, seasonality, and error components using additive decomposition models. Water quality index (WQI) derived from 17 water quality variables were used to compare Aksu upstream and Big Melen downstream water quality. Twenty-six additive decomposition models of water quality time series data including WQI had R ² values ranging from 88% for log(water temperature) (P?≤?0.001) to 3% for log(total dissolved solids) (P?≤?0.026). Linear trend models revealed that total hardness, calcium concentration, and log(nitrite concentration) had the highest rate of increase over time. Tukey’s multiple comparison pointed to significant decreases in 17 water quality variables including WQI of Big Melen downstream relative to those of Aksu upstream (P?≤?0.001). Monitoring changes in water quality on the basis of watersheds through WQI and decomposition analysis of time series data paves the way for an adaptive management process of water resources that can be tailored in response to effectiveness and dynamics of management practices.     

Hydrometeorological variables predict fecal indicator bacteria densities in freshwater: data-driven methods for variable selection

Statistical models of microbial water quality inform risk management for water recreation. Current research focuses on resource-intensive, location-specific data collection and water quality modeling, but this approach may be cost-prohibitive for risk managers responsible for numerous recreation sites. As an alternative, we tested the ability of two data-driven models, tree regression and random forests with conditional inference trees, to select readily available hydrometeorological variables for use in linear mixed effects (LME) models predicting bacterial density. The study included the Chicago Area Waterway System (CAWS) and Lake Michigan beaches and harbors in Chicago, Illinois, at which Escherichia coli and enterococci were measured seasonally in 2007–2009. Tree regression node variables reduced data dimensionality by >50 %. Variable importance ranks from random forests were used in a forward-step selection based on R ² and root mean squared prediction error (RMSPE). We found two to three variables explained bacteria densities well relative to random forests with all variables. LME models with tree- or forest-selected variables performed reasonably well (0.335?<?R ²?<?0.658). LME models for Lake Michigan had good prediction accuracy with respect to the single sample maximum standard (72–77 %), but limited sensitivity (23–62 %). Results suggest that our alternative approach is feasible and performs similarly to more resource-intensive approaches.     

固相萃取-气相色谱质谱联用法测定地表水中藻致嗅味物质

选用HLB固相萃取小柱进行富集,用二氯甲烷进行洗脱,优化了固相萃取过程,建立了固相萃取-气相色谱质谱法测定地表水中7种藻致嗅味物质的方法,浓度在0.004~0.1 mg/L范围内线性关系良好,具有较高的灵敏度和精密度,方法检出限为1×10-6~2×10-6mg/L,加标回收率为85.9%~108.6%,适用于地表水中嗅味物质的检测。     

The relative influence of nutrients and habitat on stream metabolism in agricultural streams

Stream metabolism was measured in 33 streams across a gradient of nutrient concentrations in four agricultural areas of the USA to determine the relative influence of nutrient concentrations and habitat on primary production (GPP) and respiration (CR-24). In conjunction with the stream metabolism estimates, water quality and algal biomass samples were collected, as was an assessment of habitat in the sampling reach. When data for all study areas were combined, there were no statistically significant relations between gross primary production or community respiration and any of the independent variables. However, significant regression models were developed for three study areas for GPP (r ² = 0.79–0.91) and CR-24 (r ² = 0.76–0.77). Various forms of nutrients (total phosphorus and area-weighted total nitrogen loading) were significant for predicting GPP in two study areas, with habitat variables important in seven significant models. Important physical variables included light availability, precipitation, basin area, and in-stream habitat cover. Both benthic and seston chlorophyll were not found to be important explanatory variables in any of the models; however, benthic ash-free dry weight was important in two models for GPP.     

Study on phosphorus loadings in ten natural and agricultural watersheds in subtropical region of China

Water eutrophication in subtropical regions of southern China threatens watershed health and is of major concern. However, annual phosphorus (P) loading and its dominant causes are still unclear, especially at the watershed scale. In this study, we investigated dynamic P loadings and associated factors (e.g., land use, livestock production, and runoff depth) in ten watersheds that varied in area from 9 to 5,212 ha in a hilly area of Hunan Province, China. A flowmeter was installed at the outlet of each watershed, and total P (TP) and soluble P (SP) concentrations were monitored periodically from June 2010 to October 2012. The results showed that annual P loadings (APLs) in the ten watersheds ranged from 22.8 to 247.8 kg P/km² and that P loss primarily occurred from April to June of each year during the main rainfall season in the study area. In addition, the average eutrophication (>0.05 mg P/L) ratio for stream waters was 86.7 % during the study period, which was indicative of a potentially serious condition for the local water environments. Annual P loadings were linearly related to livestock density (LD; R?=?0.92, p?<?0.01), whereas the eutrophication ratio of stream water was significantly (p?<?0.05) correlated with LD (R?=?0.61), percentage cropland (R?=?0.71), and percentage forest cover (R?=??0.68). Thus, it is concluded that the control of livestock production has the greatest potential for reducing P loadings in watersheds in this subtropical area. This will be beneficial to the amelioration and protection of local environment.     

Characteristics of atmospheric non-methane hydrocarbons during haze episode in Beijing, China

This study firstly focused on non-methane hydrocarbons (NMHCs) during three successive days with haze episode (16–18 August 2006) in Beijing. Concentrations of alkanes, alkenes, aromatic hydrocarbons, and ethyne all peaked at traffic rush hour, implying vehicular emission; and alkanes also peaked at non-traffic rush hour in the daytime, implying additional source. Especially, alkanes and aromatics clearly showed higher levels in the nighttime than that in the daytime, implying their active photochemical reactions in the daytime. Correlation coefficients (R ²) showed that propane, n-butane, i-butane, ethene, propene, and benzene correlated with ethyne (R ²?=?0.61–0.66), suggesting that their main source is vehicular emission; 2-methylpentane and n-hexane correlated with i-pentane (R ²?=?0.61–0.64), suggesting that gasoline evaporation is their main source; and ethylbezene, m-/p-xylene, and o-xylene correlated with toluene (R ²?=?0.60–0.79), suggesting that their main source is similar to that of toluene (e.g., solvent usage). The R ² of ethyne, i-pentane, and toluene with total NMHCs were 0.58, 0.76, and 0.60, respectively, indicating that ambient hydrocarbons are associated with vehicular emission, gasoline evaporation, and solvent usage. The sources of other hydrocarbons (e.g., ethane) might be natural gas leakage, biogenic emission, or long-range transport of air pollutants. Measured higher mean B/T ratio (0.78?±?0.27) was caused by the more intensive photochemical activity of toluene than benzene, still indicating the dominant emission from vehicles.     

Quantitative detection of nitrate in water and wastewater by surface-enhanced Raman spectroscopy

The presence of inorganic nitrogen species in water can be unsuitable for drinking and detrimental to the environment. In this study, a surface-enhanced Raman spectroscopy (SERS) method coupled with a commercially available gold nanosubstrate (a gold-coated silicon material) was evaluated for the detection of nitrate and nitrite in water and wastewater. Applications of SERS coupled with gold nanosubstrates resulted in an enhancement of Raman signals by a factor of ～10⁴ compared to that from Raman spectroscopy. The new method was able to detect nitrate with linear ranges of 1–10,000 mg NO₃ ^?/L (R ²?=?0.978) and 1–100 mg NO₃ ^?/L (R ²?=?0.919) for water and wastewater samples, respectively. Among the common anions, phosphate appeared to be the major interfering anion affecting nitrate measurement. Nevertheless, the percentage error of nitrate measurement in wastewater by the proposed SERS method was comparable to that by ion chromatography. The nitrate detection limits in water and wastewater samples were about 0.5 mg/L. The SERS method could simultaneously detect sulfate, which may serve as a reference standard in water. These results suggested that the SERS coupled with nanosubstrates is a promising method to determine nitrate concentrations in water and wastewater.     

Retrieval of total suspended matter (TSM) and chlorophyll-a (Chl-a) concentration from remote-sensing data for drinking water resources

The concentrations of chlorophyll-a (Chl-a) and total suspended matter (TSM) are major water quality parameters that can be retrieved using remotely sensed data. Water sampling works were conducted on 15 July 2007 and 13 September 2008 concurrent with the Indian Remote-Sensing Satellite (IRS-P6) overpass of the Shitoukoumen Reservoir. Both empirical regression and back-propagation artificial neural network (ANN) models were established to estimate Chl-a and TSM concentration with both in situ and satellite-received radiances signals. It was found that empirical models performed well on the TSM concentration estimation with better accuracy (R ² = 0.94, 0.91) than their performance on Chl-a concentration (R ² = 0.62, 0.75) with IRS-P6 imagery data, and the models accuracy marginally improved with in situ spectra data. Our results indicated that the ANN model performed better for both Chl-a (R ² = 0.91, 0.82) and TSM (R ² = 0.98, 0.94) concentration estimation through in situ collected spectra; the same trend followed for IRS-P6 imagery data (R ² = 0.75 and 0.90 for Chl-a; R ² = 0.97 and 0.95 for TSM). The relative root mean square errors (RMSEs) from the empirical model for TSM (Chl-a) were less than 15% (respectively 27.2%) with both in situ and IRS-P6 imagery data, while the RMSEs were less than 7.5% (respectively 18.4%) from the ANN model. Future work still needs to be undertaken to derive the dynamic characteristic of Shitoukoumen Reservoir water quality with remotely sensed IRS-P6 or Landsat-TM data. The algorithms developed in this study will also need to be tested and refined with more imagery data acquisitions combined with in situ spectra data.     

Study of harmful algal blooms in a eutrophic pond, Bangladesh

The purpose of this research was to analyze the underlying mechanisms and contributing factors related to the seasonal dynamic of harmful algal blooms in a shallow eutrophic pond, Bangladesh during September 2005–July 2006. Two conspicuous events were noted simultaneously throughout the study period: high concentration of phosphate–phosphorus (>3.03; SD 1.29 mg l^???1) and permanent cyanobacterial blooms {>3,981.88 × 10³ cells l^???1 (SD 508.73)}. Cyanobacterial blooms were characterized by three abundance phases, each of which was associated with different ecological processes. High nitrate–nitrogen (>2.35; SD 0.83 mg l^???1), for example, was associated with high cyanobacterial abundance, while low nitrate–nitrogen (0.36; SD 0.2 mg l^???1) was recorded during moderate abundance phase. Extremely low NO₃–N/PO₄–P ratio (>3.55, SD 2.31) was recorded, and all blooming taxa were negatively correlated with this ratio. Cyanobacterial blooms were positively correlated with temperature (r?=?0.345) and pH (0.833; p?=?0.05) and negatively correlated with transparency (r?=???0.956; p?=?0.01). Although Anabaena showed similar relationship with water quality parameters as cyanobacteria, the co-dominant Microcystis exhibited negative relationship with temperature (r?=???0.386) and nitrate–nitrogen (r?=???0.172). This was attributed to excessive growth of Anabaena that suppressed Microcystis’s growth. Planktothrix was the third most dominant taxa, while Euglena was regarded as opportunistic.     

Fate of selenium in a small urban watershed

A field study was conducted on a small urban watershed (residential and golf course dominated) in southern Nevada to assess the concentration and speciation of selenium (Se) in a series of drain lines and monitoring wells and to quantify the mass discharge of Se from the drain system. Water samples were collected on a monthly basis and analyzed for total Se, selenate (SeO₄ ⁼) and selenite (SeO₃ ⁼). In addition, where possible, flow was assessed as was, temperature, redox potential, pH, dissolved oxygen (DO) along with all major cations and anions. The data were then modeled with PhreeqC to identify selenium speciation. Results revealed a SeO₄ ⁼ dominated system with SeO₄ ⁼ concentrations ranging from 13 to 62 ppb. In the monitoring wells, 66 % of the variation in the total Se concentration could be described based on depth to groundwater, temperature and sulfate concentrations (P?<?0.001). In particular, higher total Se concentrations were predicted for shallower depth to groundwater, suggesting the solubilization of Se evapo-concentrates near the surface could be reduced by lowering water tables. The highest of all correlations was found between SeO₄ ⁼ concentrations (↑) and the sodium (↑) and DO (↑) concentrations in the monitoring wells (R ²?=?0.77, P?<?0.001). An excellent curvilinear relationship was found between total Se and the electrical conductivity in the water (R ²?=?0.73, P?<?0.001). Based on the Se data and time line identified in this study, high concentrations of Se could be expected to drain from this area for many years to come, with salinity acting as a good proxy for Se concentration. In the drain lines, Se concentrations were found to be invariant to flow (P?>?0.05). Flow discharge from the main drain system to the Las Vegas Wash was estimated at 559 acre feet during the 1 year study period. This flow was estimated to carry 4,203 Mg of salts 6.71 Mg of nitrate-N and 27.1 kg of total Se.     

Monitoring and assessment of surface water acidification following rewetting of oxidised acid sulfate soils

Large-scale exposure of acid sulfate soils during a hydrological drought in the Lower Lakes of South Australia resulted in acidification of surface water in several locations. Our aim was to describe the techniques used to monitor, assess and manage these acidification events using a field and laboratory dataset (n?=?1,208) of acidic to circum-neutral pH water samples. The median pH of the acidified (pH?<?6.5) samples was 3.8. Significant (p?<?0.05) increases in soluble metals (Al, Co, Mn, Ni and Zn above guidelines for ecosystem protection), SO₄ (from pyrite oxidation), Si (from aluminosilicate dissolution) and Ca (from carbonate dissolution and limestone addition), were observed under the acidic conditions. The log of the soluble metal concentrations, acidity and SO₄/Cl ratio increased linearly with pH. The pH, alkalinity and acidity measurements were used to inform aerial limestone dosing events to neutralise acidic water. Field measurements correlated strongly with laboratory measurements for pH, alkalinity and conductivity (r ²?≥?0.97) but only moderately with acidity (r ²?=?0.54), which could be due to difficulties in determining the indicator-based field titration endpoint. Laboratory measured acidity correlated well with calculated acidity (r ²?=?0.87, acidity present as Al^III?>>?H⁺?≈?Mn^II?>?Fe^II/III) but was about 20 % higher on average. Geochemical speciation calculations and XRD measurements indicated that solid phase minerals (schwertmannite and jarosite for Fe and jurbanite for Al) were likely controlling dissolved metal concentrations and influencing measured acidity between pH 2 and 5.     

Occurrence of haloacetic acids (HAAs) and trihalomethanes (THMs) in drinking water of Taiwan

In this study, water samples were collected from 86 water treatment plants for analysis of haloacetic acids (HAAs) and trihalomethanes (THMs) from February to March, 2007 and from July to August, 2007. Both seasonal and geographical variations of disinfection by-products (DBPs) in drinking water of Taiwan were presented. The results showed that the five HAA concentrations (HAA5) were 1.0–38.9 μg/L in the winter and 0.2–46.7 μg/L in the summer; and the total THMs were ND-99.4 μg/L in the winter and ND-133.2 μg/L in the summer. For samples taken from the main Taiwan island, dichloroacetic acid (29.4–31.7%) and trichloroacetic acid (25.3–27.6%) were the two major HAA species, and trichloromethane was the major THM species (49.9–62.2%) in finished water. For water treatment plants located on the offshore islands outside of Taiwan, high bromide concentration was found in raw water, and higher percentage of brominated THMs and HAAs were formed in the overall formation. A statistically significant (P?<?0.005) logarithmic linear regression model was found to be useful to describe the correlations between TTHM and HAA5 or nine HAAs (HAA5?=?1.219 ×TTHM ^0.754, R ²?=?0.658; HAA9?=?1.824 ×TTHM ^0.735, R ²?=?0.678). No apparent difference was observed for DBPs concentrations between finished water and distribution samples in this study.     

Urban stream syndrome in a small,lightly developed watershed: a statistical analysis of water chemistry parameters,land use patterns,and natural sources

The relationships among land use patterns, geology, soil, and major solute concentrations in stream water for eight tributaries of the Kayaderosseras Creek watershed in Saratoga County, NY, were investigated using Pearson correlation coefficients and multivariate regression analysis. Sub-watersheds corresponding to each sampling site were delineated, and land use patterns were determined for each of the eight sub-watersheds using GIS. Four land use categories (urban development, agriculture, forests, and wetlands) constituted more than 99 % of the land in the sub-watersheds. Eleven water chemistry parameters were highly and positively correlated with each other and urban development. Multivariate regression models indicated urban development was the most powerful predictor for the same eleven parameters (conductivity, TN, TP, NO $_{3}^-$ , Cl^?, HCO $_{3}^-$ , SO $_{4}^{2-}$ , Na⁺, K⁺, Ca²⁺, and Mg²⁺). Adjusted R ² values, ranging from 19 to 91 %, indicated that these models explained an average of 64 % of the variance in these 11 parameters across the samples and 70 % when Mg²⁺ was omitted. The more common R ², ranging from 29 to 92 %, averaged 68 % for these 11 parameters and 72 % when Mg²⁺ was omitted. Water quality improved most with forest coverage in stream watersheds. The strong associations between water quality variables and urban development indicated an urban source for these 11 water quality parameters at all eight sampling sites was likely, suggesting that urban stream syndrome can be detected even on a relatively small scale in a lightly developed area. Possible urban sources of Ca²⁺ and HCO $_{3}^-$ are suggested.     

Temporal variability in water quality parameters—a case study of drinking water reservoir in Florida, USA

Our objective was to evaluate changes in water quality parameters during 1983–2007 in a subtropical drinking water reservoir (area: 7 km²) located in Lake Manatee Watershed (area: 338 km²) in Florida, USA. Most water quality parameters (color, turbidity, Secchi depth, pH, EC, dissolved oxygen, total alkalinity, cations, anions, and lead) were below the Florida potable water standards. Concentrations of copper exceeded the potable water standard of <30 μg?l^?1 in about half of the samples. About 75 % of total N in lake was organic N (0.93 mg?l^?1) with the remainder (25 %) as inorganic N (NH₃-N: 0.19, NO₃-N: 0.17 mg?l^?1), while 86 % of total P was orthophosphate. Mean total N/P was <6:1 indicating N limitation in the lake. Mean monthly concentration of chlorophyll-a was much lower than the EPA water quality threshold of 20 μg?l^?1. Concentrations of total N showed significant increase from 1983 to 1994 and a decrease from 1997 to 2007. Total P showed significant increase during 1983–2007. Mean concentrations of total N (n?=?215; 1.24 mg?l^?1) were lower, and total P (n?=?286; 0.26 mg?l^?1) was much higher than the EPA numeric criteria of 1.27 mg total N l^?1 and 0.05 mg total P l^?1 for Florida’s colored lakes, respectively. Seasonal trends were observed for many water quality parameters where concentrations were typically elevated during wet months (June–September). Results suggest that reducing transport of organic N may be one potential option to protect water quality in this drinking water reservoir.     

Total suspended matter observation in the Pearl River estuary from in situ and MERIS data

Based on the cruise data collected in the Pearl River estuary (PRE) in May 2008, an empirical two-band model by using the ratio of R _rs at 629 and 671 nm was established to retrieve total suspended matter (TSM) concentration with the determination coefficient (R²) of 0.854, mean relative error (MRE) of 7.483%, and root-mean-square error (RMSE) of 1.295 mg L^???1. To match with medium resolution imaging spectrometer (MERIS) bands, in situ remote sensing reflectance was re-sampled to the bandwidth of 10 nm. The relationship between TSM and re-sampled R _rs at 620 nm (MERIS band 6) and 665 nm (MERIS band 7) are obtained (R² = 0.748, RMSE = 1.697 mg L^???1, MRE = 8.785%, n = 13). Additionally, to map the spatial distribution of TSM in the PRE, MERIS level_1B data were calibrated using a multiple linear regression model based on in situ R _rs. Another dataset collected in the PRE in January 2004 was used to validate the two-band model and also applied to map TSM distribution from MERIS image. The comparison between measured TSM values and modeled ones showed satisfactory results (R² = 0.753, MRE = 22.199%, and RMSE = 2.603 mg L^???1).     

A regional algorithm to model mesozooplankton biomass along the southwestern Bay of Bengal

A three-dimensional regression analysis attempted to model mesozooplankton (MSP) biomass using sea surface temperature (SST) and chlorophyll-a (Chl-a). The study was carried out from January 2014 to July 2015 in the southwestern Bay of Bengal (BoB) and sampling was carried out on board Sagar Manjusha and Sagar Purvi. SST ranged from 26.2 to 33.1 °C while Chl-a varied from 0.04 to 6.09 μg L^?1. During the course of the study period, there was a weak correlation (r?= 0.32) between SST and Chl-a statistically. MSP biomass varied from 0.42 to 9.63 mg C m^?3 and inversely related with SST. Two kinds of approaches were adopted to develop the model by grouping seasonal datasets (four seasonal algorithms) and comprising all datasets (one annual algorithm). Among the four functions used (linear, paraboloid, the Lorentzian and the Gaussian functions), paraboloid model was best suited. The best seasonal and annual algorithms were applied in the synchronous MODIS-derived SST and Chl-a data to estimate the MSP biomass in the southwestern BoB. The modelled MSP biomass was validated with field MSP biomass and the result was statistically significant, showing maximum regression coefficient for the seasonal algorithms (R²?=?0.60; p?=?0.627; α?= 0.05), than the annual algorithm (R²?=?0.52; p?=?0.015, α?=?0.05).