Mary Love M. Tagert

The Use of NARX Neural Networks to Forecast Daily Groundwater Levels

The lack of information to manage groundwater for irrigation is one of the biggest concerns for farmers and stakeholders in agricultural areas of Mississippi. In this study, we present a novel implementation of a nonlinear autoregressive with exogenous inputs (NARX) network to simulate daily groundwater levels at a local scale in the Mississippi River Valley Alluvial (MRVA) aquifer, located in the southeastern United States. The NARX network was trained using the Levenberg-Marquardt (LM) and Bayesian Regularization (BR) algorithms, and the results were compared to identify an optimal architecture for the forecasting of daily groundwater levels over time. The training algorithms were implemented using different hidden node combinations and delays (5, 25, 50, 75, and 100) until the optimal network was found. Eight years of daily historical input time series including precipitation and groundwater levels were used to forecast groundwater levels up to three months ahead. The comparison between LM and BR showed that NARX-BR is superior in forecasting daily levels based on the Mean Squared Error (MSE), coefficient of determination (R2), and Nash-Sutcliffe coefficient of efficiency. The results showed that BR with two hidden nodes and 100 time delays provided the most accurate prediction of groundwater levels with an error of ± 0.00119 m. This innovative study is the first of its kind and will provide significant contributions for the implementation of data-based models (DBMs) in the prediction and management of groundwater for agricultural use.

Modeling Phosphorus Loading to the Ross Barnett Reservoir Using SWAT in the Upper Pearl River Watershed in East-central Mississippi

The objective of this research was to evaluate spatially and temporally variable phosphorus loading to the Ross Barnet Reservoir in east-central Mississippi using a modeling approach. Modeling methods were developed to model livestock, poultry, and human sources of nutrients from the Upper Pearl River watershed (UPRW-7,885 km2).The Soil and Water Assessment Tool (SWAT) model was applied to evaluate average monthly flow, sediment, total nitrogen (N), and total phosphorus (P) loading to the Ross Barnett Reservoir inlet.

Evaluation of Seasonally Classified Inputs for the Prediction of Daily Groundwater Levels: NARX Networks Vs Support Vector Machines

Farmers and stakeholders who use groundwater for irrigation need efficient and cost-effective tools to support sustainable crop production. The variation of groundwater levels at local scale and its continuous use for agriculture intensifies the demand for reliable groundwater information. However, groundwater levels are very dynamic and difficult to predict under traditional modeling approaches, and manual monitoring of wells is costly and time-consuming. Simplified but powerful machine learning models represent a practical alternative to support groundwater management decisions at the farm scale. The predictive capacity of a nonlinear autoregressive with exogenous inputs (NARX) artificial neural network (ANN) and a support vector regression (SVR) trained with a radial basis function (RBF) algorithm were evaluated for an irrigation well located in a highly productive agricultural region in the southeastern USA. We used separately multiple years of daily historical time series classified by summer (withdrawal) and winter (recharge) seasons and evaluated the impacts of this division in the models’ predictive capability. Results showed that SVR had a better modeling performance based on the mean squared error (MSE) and prediction trend for both seasons. In addition, our study suggests that the prediction of daily levels with input time series classified by seasons provides higher accuracy than using the entire withdrawal and recharge periods as a whole. Results also indicate that the recharge season becomes a linear problem, which substantially reduces the SVR modeling computational requirements. The application of our proposed modeling approach in the management of groundwater sources for irrigation provides important information, at short time scale, for the estimation of groundwater variability at local scale.

Water Quality Dynamics in Agricultural Ponds in Mississippi: In situ Measured Parameters

Abstract. The intensification of agricultural production in northwestern Mississippi over the past few decades has resulted in an increase in water withdrawals for irrigation and a dramatic decline in groundwater levels in the Mississippi River Valley Alluvial (MRVA) aquifer. Across the Mississippi Delta Region (MDR), groundwater is the primary and ultimate source of water for irrigation of crops when other surface water sources are absent or have been depleted. The MDR represents 93.7% of the total irrigated area in Mississippi, of which 97.4% withdraws groundwater from the MRVA aquifer at an average rate of roughly 530 m 3 yr -1 for each hectare of cultivated land. The increased demand on groundwater is of special concern for regulatory agencies, the scientific community, and producers because such withdrawal rates exceed the annual recharge capacity of the aquifer. Multi-purpose best management practices (BMPs) as on-farm water storage (OFWS) systems have been implemented throughout the MDR in an attempt to reduce groundwater withdrawals, offer further sources for irrigation, improve water quality downstream, and promote water conservation. An OFWS system typically includes a tailwater recovery ditch that collects surface runoff and irrigation tailwater and an agricultural pond, or on-farm reservoir, to store surface water for later use. While ponds provide stored surface water for irrigation, there is a lack of field data to better understand the water quality of this source, and the implications of using this water on agriculture is scarce in the scientific literature. This study presents preliminary results of monitoring and analysis of in-situ water quality parameters measured in two agricultural ponds established in two farms located in Porter Bayou Watershed, Mississippi. Water supply from the investigated ponds is suitable for irrigation of crops although the preliminary results so far are for only three water quality parameters that are of central interest to irrigated agriculture.

Effects of Hydroclimate on In-ditch Water Quality: Case Study of Two Tailwater Recovery Ditches in Mississippi

Abstract. Nutrient loss via soil erosion caused by surface and irrigation runoff is of great concern for agricultural producers and the scientific community. Nutrients reach water bodies and in many instances exceed acceptable levels, which can adversely impact aquatic ecosystems. While best management practices (BMPs) are implemented with the goal of controlling nutrient pollution in agricultural landscapes, their efficiencies are highly sensitive to local changes in hydroclimate and irrigation schemes. This paper analyzes the impact of rainfall characteristics on water quality in two tailwater recovery (TWR) ditches implemented at two farms in the Porter Bayou watershed, Mississippi. Forty-seven rainfall events were selected to match water quality monitoring of the ditches conducted from March 2012 to March 2016. These events were classified according to their intensities and durations into five classes and then correlated with in-ditch nitrate nitrogen (NO 3 – N) concentrations. Rainfall frequency ranked from highest to lowest was class II > class III > class I = class V > class IV. In terms of total rainfall depth, the following order was class V > class III > class II > class IV > class I. NO 3 – N concentration correlated negatively with time between previous rainfall and sampling events and time before next-to-last rainfall event, and positively with depth of next-to-last rainfall event, which suggests that the possible nutrient reduction ability of the TWR ditches could be overwhelmed by high-magnitude intensity and recurrent rainfall events. According to our study, rainfall events classified as II and III were the most frequent and also the most occurred to trigger higher concentrations of NO 3 – N in the investigated ditches. Preliminary results of the present study will contribute to the improvement of nutrient loss management in agricultural landscapes in Mississippi and enhance understanding of the BMP‘s responses to agro-hydrologic variability.

On-Farm Water Storage Systems in Porter Bayou Watershed, Mississippi

Since the 1970’s, groundwater levels in the Mississippi Alluvial Aquifer have decreased at a rate of approximately 100,000 acre-feet per year due to increased irrigated acres. There are roughly 13,000 permitted irrigation wells dependent on water from the Mississippi Alluvial Aquifer. Farmers and landowners are faced with two major issues with regard to sustainably managing agroecosystems in the Mississippi Delta region, namely, the declining groundwater levels in the Mississippi Delta Shallow Alluvial Aquifer, and nutrient loads into the Mississippi River and the Gulf of Mexico. Due to concerns over groundwater declines and increasing fuel costs to run irrigation pumps, farmers in the Mississippi Delta region have begun implementing irrigation conservation measures, such as tailwater recovery (TWR) ditches and storage ponds to capture irrigation and surface water runoff from the field for later use. Monitoring of two on-farm water storage (OFWS) systems at Pitts and Metcalf farms in Porter Bayou Water shed was initiated in February 2012. Water samples from storage ponds and TWR ditches were analyzed for different water quality parameters. Additional information such as water level in the TWR ditch and water weather station were recorded at 10-minute and 15 minute intervals. An inventory of installed and pending OFWS systems in Porter Bayou Watershed will be conducted as part of this study.