Dynamic factor analysis of groundwater quality trends in an agricultural area adjacent to Everglades National Park

Dynamic multivariate analysis for pollution assessment and river habitat conservation in the Vietnamese La Buong watershed

Analysis of temporal patterns of high-dimensional time-series water quality data is essential for pollution management worldwide. This study has applied dynamic factor analysis (DFA) and cluster analysis (CA) to analyze time-series water quality data monitored at the five stations installed along the La Buong river in Southern Vietnam. Application of the DFA identified two types of temporal patterns, one of the run-off driven parameters (total suspended solid (TSS), turbidity, and iron) and the other of diffuse source pollution. The association of the variables like BOD₅ and COD at most stations to the run-off-driven parameters revealed their sharing of drivers. On the contrary, separating variables like phosphate (PO₄³) at the three upstream stations from the run-off patterns suggested their local point-source origin. The DFA-derived factors were later used in the time-point CA to explore the seasonality of water quality parameters and their pollution intensities compared to regulatory levels. The result suggested intensification in wet season of Fe, TSS, BOD₅, and COD concentrations at most sites, which are unobservable in run-off detached parameters like reactive nitrogen, phosphate (PO₄^3-), and E. coli. These findings generated robust insights to support water quality management for river habitat conservation.

A New Approach in Determining the Decadal Common Trends in the Groundwater Table of the Watershed of Lake “Neusiedlersee”

Shallow groundwater is one of the primary sources of fresh water, providing river base-flow and root-zone soil water between precipitation events. However, with urbanization and the increase in demand for water for irrigation, shallow groundwater bodies are being endangered. In the present study, 101 hydrographs of shallow groundwater monitoring wells from the watershed of the westernmost brackish lake in Europe were examined for the years 1997–2012 using a combination of dynamic factor and cluster analyses. The aims were (i) the determination of the main driving factors of the water table, (ii) the determination of the spatial distribution and importance of these factors, and (iii) the estimation of shallow groundwater levels using the obtained model. Results indicate that the dynamic factor models were capable of accurately estimating the hydrographs (avg. mean squared error = 0.29 for standardized water levels), meaning that the two driving factors identified (evapotranspiration and precipitation) describe most of the variances of the fluctuations in water level. Both meteorological parameters correlated with an obtained dynamic factor (r = −0.41 for evapotranspiration & r = 0.76 for precipitation). The strength of these effects displayed a spatial pattern, as did the factor loadings. On this basis, the monitoring wells could be objectively distinguished into two groups using hierarchical cluster analysis and verified by linear discriminant analysis in 98% of the cases. This grouping in turn was determined to be primarily related to the elevation and the geology of the area. It can be concluded that the application of the data analysis toolset suggested herein permits a more efficient, objective, and reproducible delineation of the primary driving factors of the shallow groundwater table in the area. Additionally, it represents an effective toolset for the forecasting of water table variations, a quality which, in the view of the likelihood of further climate change to come, is a distinctive advantage. The knowledge of these factors is crucial to a better understanding of the hydrogeological processes that characterize the water table and, thus, to developing a proper water resource management strategy for the area.

The Role of Environmental Background Processes in Determining Groundwater Level Variability—An Investigation of a Record Flood Event Using Dynamic Factor Analysis

Since groundwater is a major source of water for drinking and for industrial and irrigation uses, the identification of the environmental processes determining groundwater level fluctuation is potentially a matter of great consequence, especially in light of the fact that the frequency of extreme climate events may be expected to increase, causing changes in groundwater recharge systems. In the recent study, data measured at a frequency of one hour were collected from the Szigetköz, an inland delta of the Danube. These were then used to determine the presence, or not, and magnitude of any hidden environmental background factors that may be causing groundwater level fluctuations. Through the application of dynamic factor analysis, it was revealed that changes in groundwater level are mainly determined by (i) the water level of neighboring rivers and (ii) evapotranspiration. The intensity of these factors may also be estimated spatially. If the background factors determined by dynamic factor analysis do indeed figure in the linear model as variables, then the time series of groundwater levels can be said to have been accurately estimated with the use of linear regression. The accuracy of the estimate is indicated by the fact that adjusted coefficient of determination exceeds 0.9 in 80% of the wells. The results, via an enhanced understanding of the reasons for changes in the fluctuation of groundwater, could assist in the development of sustainable water management and irrigation strategies and the preparation for varying potential climate change scenarios.

The Latitudes, Attitudes, and Platitudes of Watershed Phosphorus Management in North America

Phosphorus (P) plays a crucial role in agriculture as a primary fertilizer nutrient-and as a cause of the eutrophication of surface waters. Despite decades of efforts to keep P on agricultural fields and reduce losses to waterways, frequent algal blooms persist, triggering not only ecological disruption but also economic, social, and political consequences. We investigate historical and persistent factors affecting agricultural P mitigation in a transect of major watersheds across North America: Lake Winnipeg, Lake Erie, the Chesapeake Bay, and Lake Okeechobee/Everglades. These water bodies span 26 degrees of latitude, from the cold climate of central Canada to the subtropics of the southeastern United States. These water bodies and their associated watersheds have tracked trajectories of P mitigation that manifest remarkable similarities, and all have faced challenges in the application of science to agricultural management that continue to this day. An evolution of knowledge and experience in watershed P mitigation calls into question uniform solutions as well as efforts to transfer strategies from other arenas. As a result, there is a need to admit to shortcomings of past approaches, plotting a future for watershed P mitigation that accepts the sometimes two-sided nature of Hennig Brandt's "Devil's Element."

Highway paving in the southwestern Amazon alters long-term trends and drivers of regional vegetation dynamics

Infrastructure development, specifically road paving, contributes socio-economic benefits to society worldwide. However, detrimental environmental effects of road paving have been documented, most notably increased deforestation. Beyond deforestation, we hypothesize that road paving introduces “unseen” regional scale effects on forests, due to changes to vegetation dynamics. To test this hypothesis, we focus on the tri-national frontier in the southwestern Amazon that has been subject to construction of the Inter-Oceanic Highway (IOH) between 1987 and 2010. We use a long-term remotely sensed vegetation index as a proxy for vegetation dynamics and combine these with field-based socio-ecological data and biophysical data from global datasets. We find 4 areas of shared vegetation dynamics associated with increasing extent of road paving. Applying Dynamic Factor Analysis, an exploratory dimension-reduction time series analysis technique, we identify common trends and covariates in each area. Common trends, indicating underlying unexplained effects, become relatively less important as paving increases, and covariates increase in importance. The common trends are dominated by lower frequency signals possibly embodying long-term climate variability. Human-related covariates become more important in explaining vegetation dynamics as road paving extent increases, particularly family density and travel time to market. Natural covariates such as minimum temperature and soil moisture become less important. The change in vegetation dynamics identified in this study indicates a possible change in ecosystem services along the disturbance gradient. While this study does not include all potential factors controlling dynamics and disturbance of vegetation in the region, it offers important insights for management and mitigation of effects of road paving projects. Infrastructure planning initiatives should make provisions for more detailed vegetation monitoring after road completion, with a broader focus than just deforestation. The study highlights the need to mitigate population-driven pressures on vegetation like family density and access to new markets.

Phytoplankton dynamics of a subtropical reservoir controlled by the complex interplay among hydrological, abiotic, and biotic variables

This study was conducted to identify the key factors related to the spatiotemporal variations in phytoplankton abundance in a subtropical reservoir from 2006 to 2010 and to assist in developing strategies for water quality management. Dynamic factor analysis (DFA), a dimension-reduction technique, was used to identify interactions between explanatory variables (i.e., environmental variables) and abundance (biovolume) of predominant phytoplankton classes. The optimal DFA model significantly described the dynamic changes in abundances of predominant phytoplankton groups (including dinoflagellates, diatoms, and green algae) at five monitoring sites. Water temperature, electrical conductivity, water level, nutrients (total phosphorus, NO₃-N, and NH₃-N), macro-zooplankton, and zooplankton were the key factors affecting the dynamics of aforementioned phytoplankton. Therefore, transformations of nutrients and reactions between water quality variables and aforementioned processes altered by hydrological conditions may also control the abundance dynamics of phytoplankton, which may represent common trends in the DFA model. The meandering shape of Shihmen Reservoir and its surrounding rivers caused a complex interplay between hydrological conditions and abiotic and biotic variables, resulting in phytoplankton abundance that could not be estimated using certain variables. Additional water quality and hydrological variables at surrounding rivers and monitoring plans should be executed a few days before and after reservoir operations and heavy storm, which would assist in developing site-specific preventive strategies to control phytoplankton abundance.

A time series analysis of multiple ambient pollutants to investigate the underlying air pollution dynamics and interactions

Understanding the temporal dynamics and interactions of particulate matter (PM) concentration and composition is important for air quality control. This paper applied a dynamic factor analysis method (DFA) to reveal the underlying mechanisms of nonstationary variations in twelve ambient concentrations of aerosols and gaseous pollutants, and the associations with meteorological factors. This approach can consider the uncertainties and temporal dependences of time series data. The common trends of the yearlong and three selected diurnal variations were obtained to characterize the dominant processes occurring in general and specific scenarios in Taipei during 2009 (i.e., during Asian dust storm (ADS) events, rainfall, and under normal conditions). The results revealed the two distinct yearlong NOx transformation processes, and demonstrated that traffic emissions and photochemical reactions both critically influence diurnal variation, depending upon meteorological conditions. During an ADS event, transboundary transport and distinct weather conditions both influenced the temporal pattern of identified common trends. This study shows the DFA method can effectively extract meaningful latent processes of time series data and provide insights of the dominant associations and interactions in the complex air pollution processes.

Groundwater salinity in a floodplain forest impacted by saltwater intrusion

Coastal wetlands occupy a delicate position at the intersection of fresh and saline waters. Changing climate and watershed hydrology can lead to saltwater intrusion into historically freshwater systems, causing plant mortality and loss of freshwater habitat. Understanding the hydrological functioning of tidally influenced floodplain forests is essential for advancing ecosystem protection and restoration goals, however finding direct relationships between hydrological inputs and floodplain hydrology is complicated by interactions between surface water, groundwater, and atmospheric fluxes in variably saturated soils with heterogeneous vegetation and topography. Thus, an alternative method for identifying common trends and causal factors is required. Dynamic factor analysis (DFA), a time series dimension reduction technique, models temporal variation in observed data as linear combinations of common trends, which represent unexplained common variability, and explanatory variables. DFA was applied to model shallow groundwater salinity in the forested floodplain wetlands of the Loxahatchee River (Florida, USA), where altered watershed hydrology has led to changing hydroperiod and salinity regimes and undesired vegetative changes. Long-term, high-resolution groundwater salinity datasets revealed dynamics over seasonal and yearly time periods as well as over tidal cycles and storm events. DFA identified shared trends among salinity time series and a full dynamic factor model simulated observed series well (overall coefficient of efficiency, Ceff=0.85; 0.52≤Ceff≤0.99). A reduced multilinear model based solely on explanatory variables identified in the DFA had fair to good results (Ceff=0.58; 0.38≤Ceff≤0.75) and may be used to assess the effects of restoration and management scenarios on shallow groundwater salinity in the Loxahatchee River floodplain.

Beyond precipitation: physiographic gradients dictate the relative importance of environmental drivers on Savanna vegetation

BACKGROUND

Understanding the drivers of large-scale vegetation change is critical to managing landscapes and key to predicting how projected climate and land use changes will affect regional vegetation patterns. This study aimed to improve our understanding of the role, magnitude and spatial distribution of the key environmental factors driving vegetation change in southern African savanna, and how they vary across physiographic gradients.

METHODOLOGY/PRINCIPAL FINDINGS

We applied Dynamic Factor Analysis (DFA), a multivariate times series dimension reduction technique to ten years of monthly remote sensing data (MODIS-derived normalized difference vegetation index, NDVI) and a suite of environmental covariates: precipitation, mean and maximum temperature, soil moisture, relative humidity, fire and potential evapotranspiration. Monthly NDVI was described by cyclic seasonal variation with distinct spatiotemporal patterns in different physiographic regions. Results support existing work emphasizing the importance of precipitation, soil moisture and fire on NDVI, but also reveal overlooked effects of temperature and evapotranspiration, particularly in regions with higher mean annual precipitation. Critically, spatial distributions of the weights of environmental covariates point to a transition in the importance of precipitation and soil moisture (strongest in grass-dominated regions with precipitation<750 mm) to fire, potential evapotranspiration, and temperature (strongest in tree-dominated regions with precipitation>950 mm).

CONCLUSIONS/SIGNIFICANCE

We quantified the combined spatiotemporal effects of an available suite of environmental drivers on NDVI across a large and diverse savanna region. The analysis supports known drivers of savanna vegetation but also uncovers important roles of temperature and evapotranspiration. Results highlight the utility of applying the DFA approach to remote sensing products for regional analyses of landscape change in the context of global environmental change. With the dramatic increase in global change research, this methodology augurs well for further development and application of spatially explicit time series modeling to studies at the intersection of ecology and remote sensing.

Investigation of long-term trends in selected physical and chemical parameters of inflows to Everglades National Park, 1977-2005

Data of seven water-quality parameters from inflows to the Everglades National Park were collected at three monitoring stations and analyzed for temporal trends. The best-fit models for the existence of trends were evaluated. The Kolmogorov-Smirnov test was used to select the theoretical distribution which best fit the data. Simple regression was used to examine the parameters for concentration-discharge relationships. The power and linear models were found to better describe the concentration-discharge relationships. Loess trend lines indicated a similar trend period of color value change during the selected period at three stations. The sharp decrease in color after 1990 at each station is consistent with the beneficial impacts of control measures, which include Best Management Practices implementation in the Everglades Agricultural Area, water management improvement, and the construction of additional stormwater treatment areas. The existence of trend analysis was performed by using the uncensored seasonal Kendall test. Conductivity and color decreased significantly at two (S12A and S333) of three stations. Alkalinity decreased significantly at S333. A "best-fit" model was selected to describe a trend change with statistical significance; the second-order equation provides a better description of the trend. This study also indicates that by using the routinely measured water-quality parameters, it may be easier to quantify the changes in water quality to aid in making water resources management decisions.

Water quality trends at inflows to Everglades National Park, 1977-2005

Restoration of the Florida Everglades is important for the health of the natural system, including both the "River of Grass" and its downstream estuaries. Water quality improvement is one indicator of successful restoration in this complex ecosystem. Using the period of record of 1977 through 2005, we evaluated data from seven inflow sites to the Everglades National Park (ENP) for temporal trends of various forms of phosphorus (P) and nitrogen (N) and analyzed them using principal component analysis and factor analysis without flow adjustments. Locally estimated scatter plot smoothing (LOESS) trend lines identified two inflection points (three time periods) of changing trend in total P (TP) concentration at the seven sites. Results indicated that overall water quality in ENP inflow improved from 1977 to 2005, with significant downward trends in TP concentration. The overall trend ofTP is probably mediated by hydrology, which is evident by a negative relationship between flow and annual average TP concentration at the majority of stations within the available data, although additional changes in vegetation due to hydroperiod may have some effects. Total N (TN), total Kjeldahl N, and total organic N concentrations also generally decreased at inflow sites. Water quality standards for TP, TN, and NH4+ -N were exceeded at selected sites during the study period. Principle component analysis and factor analysis detected a grouping of sampling sites related to the water delivery system that could be used as indicators to better manage monitoring resources. Study results suggest that water quality data analyses could provide additional insight into the success of a restoration management plan and on how monitoring may be modified for more efficient use ofresources.

Effects of rainfall on water quality in six sequentially disposed fishponds with continuous water flow

An investigation was carried out during the rainy period in six semi-intensive production fish ponds in which water flowed from one pond to another without undergoing any treatment. Eight sampling sites were assigned at pond outlets during the rainy period (December-February). Lowest and highest physical and chemical parameters of water occurred in pond P1 (a site near the springs) and in pond P4 (a critical site that received allochthonous material from the other ponds and also from frog culture ponds), respectively. Pond sequential layout caused concentration of nutrients, chlorophyll-a and conductivity. Seasonal rains increased the water flow in the ponds and, consequently, silted more particles and other dissolved material from one fish pond to another. Silting increased limnological variables from P3 to P6. Although results suggest that during the period under analysis, rainfall affected positively the ponds' water quality and since the analyzed systems have been aligned in a sequential layout with constant water flow from fish ponds and parallel tanks without any previous treatment, care has to be taken so that an increase in rain-induced water flow does not have a contrary effect in the fish ponds investigated.

Dynamic factor analysis of groundwater quality trends in an agricultural area adjacent to Everglades National Park

The extensive eastern boundary of Everglades National Park (ENP) in south Florida (USA) is subject to one of the most expensive and ambitious environmental restoration projects in history. Understanding and predicting the water quality interactions between the shallow aquifer and surface water is a key component in meeting current environmental regulations and fine-tuning ENP wetland restoration while still maintaining flood protection for the adjacent developed areas. Dynamic factor analysis (DFA), a recent technique for the study of multivariate non-stationary time-series, was applied to study fluctuations in groundwater quality in the area. More than two years of hydrological and water quality time series (rainfall; water table depth; and soil, ground and surface water concentrations of N-NO3-, N-NH4+, P-PO4(3-), Total P, F-and Cl-) from a small agricultural watershed adjacent to the ENP were selected for the study. The unexplained variability required for determining the concentration of each chemical in the 16 wells was greatly reduced by including in the analysis some of the observed time series as explanatory variables (rainfall, water table depth, and soil and canal water chemical concentration). DFA results showed that groundwater concentration of three of the agrochemical species studied (N-NO3-, P-PO4(3-)and Total P) were affected by the same explanatory variables (water table depth, enriched topsoil, and occurrence of a leaching rainfall event, in order of decreasing relative importance). This indicates that leaching by rainfall is the main mechanism explaining concentration peaks in groundwater. In the case of N-NH4+, in addition to leaching, groundwater concentration is governed by lateral exchange with canals. F-and Cl- are mainly affected by periods of dilution by rainfall recharge, and by exchange with the canals. The unstructured nature of the common trends found suggests that these are related to the complex spatially and temporally varying land use patterns in the watershed. The results indicate that peak concentrations of agrochemicals in groundwater could be reduced by improving fertilization practices (by splitting and modifying timing of applications) and by operating the regional canal system to maintain the water table low, especially during the rainy periods.