Zhulu Lin

Assessing secondary soil salinization risk based on the PSR sustainability framework

Risk assessment of secondary soil salinization, which is caused in part by the way people manage the land, is an essential challenge to agricultural sustainability. The objective of our study was to develop a soil salinity risk assessment methodology by selecting a consistent set of risk factors based on the conceptual Pressure-State-Response (PSR) sustainability framework and incorporating the grey relational analysis and the Analytic Hierarchy Process methods. The proposed salinity risk assessment methodology was demonstrated through a case study of developing composite risk index maps for the Yinchuan Plain, a major irrigation agriculture district in northwest China. Fourteen risk factors were selected in terms of the three PSR criteria: pressure, state, and response. The results showed that the salinity risk in the Yinchuan Plain was strongly influenced by the subsoil and groundwater salinity, land use, distance to irrigation canals, and depth to groundwater. To maintain agricultural sustainability in the Yinchuan Plain, a suite of remedial and preventative actions were proposed to manage soil salinity risk in the regions that are affected by salinity at different levels and by different salinization processes. The weight sensitivity analysis results also showed that the overall salinity risk of the Yinchuan Plain would increase or decrease as the weights for pressure or response risk factors increased, signifying the importance of human activities on secondary soil salinization. Ideally, the proposed methodology will help us develop more consistent management tools for risk assessment and management and for control of secondary soil salinization.

Regional land salinization assessment and simulation through cellular automaton-Markov modeling and spatial pattern analysis

Land salinization and desalinization are complex processes affected by both biophysical and human-induced driving factors. Conventional approaches of land salinization assessment and simulation are either too time consuming or focus only on biophysical factors. The cellular automaton (CA)-Markov model, when coupled with spatial pattern analysis, is well suited for regional assessments and simulations of salt-affected landscapes since both biophysical and socioeconomic data can be efficiently incorporated into a geographic information system framework. Our hypothesis set forth that the CA-Markov model can serve as an alternative tool for regional assessment and simulation of land salinization or desalinization. Our results suggest that the CA-Markov model, when incorporating biophysical and human-induced factors, performs better than the model which did not account for these factors when simulating the salt-affected landscape of the Yinchuan Plain (China) in 2009. In general, the CA-Markov model is best suited for short-term simulations and the performance of the CA-Markov model is largely determined by the availability of high-quality, high-resolution socioeconomic data. The coupling of the CA-Markov model with spatial pattern analysis provides an improved understanding of spatial and temporal variations of salt-affected landscape changes and an option to test different soil management scenarios for salinity management.

Modeling Phosphorus in the Lake Allatoona Watershed Using SWAT: II. Effect of Land Use Change

Lake Allatoona is a large reservoir northeast of metropolitan Atlanta, GA, threatened by excessive algal growth. We used the calibrated Soil and Water Assessment Tool (SWAT) models developed in our companion paper to estimate the annual P load to Lake Allatoona in 1992 and in 2001 after significant changes occurred in land use. Land use data in 1992 and 2001 from the Multi-Resolution Land Characteristics (MRLC) Consortium showed that forest land use decreased during this period by about 20%, urban land use increased by about 225%, and pasture land uses increased by about 50%. Simulation results showed that the P load to Lake Allatoona increased from 176.5 to 207.3 Mg, which were 87.8% and 103.1%, respectively, of the total P (TP) annual cap (201 Mg) set by the Georgia Environmental Protection Division (GAEPD) for discharge into Lake Allatoona. In the early 1990s, the greatest sources of the TP load to Lake Allatoona (and their percentages of the total load) were pasture (33.6%), forest (27.5%), and point sources (25.0%). Urban land uses contributed about 6.0% and row-crop agriculture contributed about 6.8%. A decade later, the greatest two TP sources were pasture (52.7%) and urban (20.9%) land uses. Point-source P loads decreased significantly to 11.6%. Permit limits on poultry processing plants reduced the point-source P loads, but increasing urban and pasture land uses increased nonpoint sources of P. To achieve further reductions in the P load to Lake Allatoona, contributions from pasture and urban nonpoint sources will need to be addressed.

Accounting for structural error and uncertainty in a model: An approach based on model parameters as stochastic processes

The significance of model structure error and uncertainty (MSEU), sometimes referred to as conceptual error, is rarely adequately recognized. MSEU, moreover, is not an esoteric matter of little consequence to the formation of policy for environmental protection and ameliorating the prospective effects of climate change. The paper presents an approach to accounting for MSEU in which the parameters of a model are treated as stochastic processes and modeled as Generalized Random Walks. Our approach is inspired by the algorithms of recursive estimation and filtering theory. In particular, given an innovations representation of the models structure, we are able to exploit the dichotomy of what is considered to be the {presumed known} in the models structure and its complement, the {acknowledged unknown}. Two conceptually different groups of model parameters attach to this dichotomy: those familiar to us as the conventional parameters in a models structure; and those having to do with the way in which past (systematic) forecasting errors - in fact, the innovations errors - are distributed (fed back) into the generation of future predictions through a gain matrix (in the sense of filtering theory). A hypothetical biological system with nonlinear dynamics is specified as the prototypical case study for assessing and comparing the performance of our proposed approach with three other approaches to accounting for MSEU: model fitting error; the expansion of parametric uncertainty; and Bayesian model averaging. Our predictive test cases are constructed around future conditions in which the pattern of input disturbances of the systems behavior is broadly similar to that of their past observed pattern (as used for prior identification, or calibration, of the model). In specific terms, however, future input disturbance patterns are significantly different. Our new approach and that of Bayesian model averaging are found to perform well on this hypothetical system; the performances of the approaches of model fitting error and the expansion of parametric uncertainty are shown to be inferior.

Modeling Phosphorus in the Lake Allatoona Watershed Using SWAT: I. Developing Phosphorus Parameter Values

Lake Allatoona is a large reservoir north of Atlanta, GA, that drains an area of about 2870 km2 scheduled for a phosphorus (P) total maximum daily load (TMDL). The Soil and Water Assessment Tool (SWAT) model has been widely used for watershed-scale modeling of P, but there is little guidance on how to estimate P-related parameters, especially those related to in-stream P processes. In this paper, methods are demonstrated to individually estimate SWAT soil-related P parameters and to collectively estimate P parameters related to stream processes. Stream related parameters were obtained using the nutrient uptake length concept. In a manner similar to experiments conducted by stream ecologists, a small point source is simulated in a headwater sub-basin of the SWAT models, then the in-stream parameter values are adjusted collectively to get an uptake length of P similar to the values measured in the streams in the region. After adjusting the in-stream parameters, the P uptake length estimated in the simulations ranged from 53 to 149 km compared to uptake lengths measured by ecologists in the region of 11 to 85 km. Once the a priori P-related parameter set was developed, the SWAT models of main tributaries to Lake Allatoona were calibrated for daily transport. Models using SWAT P parameters derived from the methods in this paper outperformed models using default parameter values when predicting total P (TP) concentrations in streams during storm events and TP annual loads to Lake Allatoona.

A biogeochemical model for metabolism and nutrient cycling in a Southeastern Piedmont impoundment

Abstract While non-point nutrient loads are important determinants of biological productivity in Southeastern Piedmont impoundments, productivity can be attenuated by concomitant sediment loads that reduce the biological availability of these nutrients. A biogeochemical model is proposed that explicitly accounts for the effects of sediment–nutrient interactions on multiple components of phytoplankton metabolism dynamics, including algal photosynthesis and respiration, pH, carbonate speciation, dissolved oxygen, and biochemical oxygen demand. Sediment–nutrient interactions relate nutrient uptake and release to pH, sediment oxygen demand, sediment organic matter, and iron. pH is a state variable in our model, affects sediment–nutrient adsorption, and constrains model parameters. The model replicates water quality observations in a small Southeastern Piedmont impoundment and suggests that pH-dependent sediment–nutrient adsorption dominates both orthophosphate and ammonium dynamics, with phosphate adsorption being controlled by ligand exchange to iron oxides, and ammonium adsorption being controlled by the cation exchange capacity. Sediment organic matter accumulation and decay also affects nutrient availability, and may explain the long-term increase of hypolimnetic dissolved oxygen deficit in Lake Lanier, a large Southeastern Piedmont impoundment.

Subsurface Drainage and Subirrigation Effects on Water Quality in Southeast North Dakota

Rising water tables, increased soil salinity, and poor trafficability have prompted rapid expansion of subsurface drainage in the Red River Valley of the North in eastern North Dakota and northwestern Minnesota. A conventional subsurface drainage (CD) and subirrigation (SI) field study was conducted in southeast North Dakota from 2008 to 2010 to investigate drainage and subirrigation effects on water quality. Water samples were collected biweekly from a sump pump structure (used as the water inlet and outlet) and 16 observation wells within the field. Water quality variables included chloride (Cl-), electrical conductivity (EC), total dissolved solids (TDS), sodium adsorption ratio (SAR), sodium (Na+), orthophosphate (PO4-P), ammonium (NH4-N), nitrite and nitrate (NOx-N), Kjeldahl nitrogen (TKN), and total nitrogen (TN). A three-factor partially nested design and a general linear model with random effects were employed to compare the effects of water management treatment, distance to drain, and well locations (soil heterogeneity) on water quality. The most significant water quality difference was found at the outlet structure, where a significant difference (p < 0.001) between the CD and SI water was found for all ten variables. The water quality of the drainage water was better than the subirrigation water from the aquifer, except for the NOx-N, EC, and TDS concentrations. Well water Cl- concentrations inside the field were significantly greater in SI compared with CD water; EC, TDS, SAR, and Na+ were not. In contrast, EC, TDS, SAR, and Na+ were significantly higher at two well locations, indicating that soil heterogeneity affected the water quality. Due to SI practice, a significant difference for Cl-, SAR, and Na+ was found between the locations closest to and farthest from the drains during the SI practice, which implies that the SI process may cause soil properties to change in the future. Overall, well locations significantly affected PO4-P, NOx-N, and TN, indicating that the soil physical and chemical properties affected the water quality, and these effects could overcome the difference due to different water treatments.

Incremental infrastructure transitions towards cities as forces for good in the environment

As source-separation has been proposed as a sustainable alternative to the current wastewater treatment strategy, the improvement in sustainability associated with such transition needs to be evaluated. However, given the difficulties and uncertainty in both technological and social aspects, such transition will most probably be happening gradually, step by step. Here with a computational case study based on the city of Atlanta within the Upper Chattahoochee watershed in the southeastern United States, we simulated three steps of transition, i.e. 50%, 70%, and 100% ANS-separation (Anthropogenic Nutrient Solution) is reached. The economic sustainability of these transition strategies was evaluated with total annual economic cost (TAEC), and the environmental sustainability was evaluated with three indicators, i.e. ecological footprints (EF), flux of materials passing through the city in its context of global material cycles, and pulse rate in terms of the spectrum of disturbance frequency to which the city is subject. The simulation results showed that compared with the current strategy, the ANS-separation has significantly lower TAEC, lower EF, lower pollutant discharge, higher recovery of nutrient and energy, and more beneficial manipulation of perturbation regimes of the citys environment. These advantages increase with the rate of ANS-separation.

Identifying and managing risk factors for salt-affected soils: a case study in a semi-arid region in China.

Soil salinization and desalinization are complex processes caused by natural conditions and human-induced risk factors. Conventional salinity risk identification and management methods have limitations in spatial data analysis and often provide an inadequate description of the problem. The objectives of this study were to identify controllable risk factors, to provide response measures, and to design management strategies for salt-affected soils. We proposed to integrate spatial autoregressive (SAR) model, multi-attribute decision making (MADM), and analytic hierarchy process (AHP) for these purposes. Our proposed method was demonstrated through a case study of managing soil salinization in a semi-arid region in China. The results clearly indicated that the SAR model is superior to the OLS model in terms of risk factor identification. These factors include groundwater salinity, paddy area, corn area, aquaculture (i.e., ponds and lakes) area, distance to drainage ditches and irrigation channels, organic fertilizer input, and cropping index, among which the factors related to human land use activities are dominant risk factors that drive the soil salinization processes. We also showed that ecological irrigation and sustainable land use are acceptable strategies for soil salinity management.

Model Structure Identification and the Growth of Knowledge

This chapter honors Peter, then, in recounting my career-long experience (1970–2010) of staring down the devilishly difficult: the problem of model structure identification—of using models for discovery. I still regard this matter as one of the grand challenges of environmental modeling (Beck et al., White Paper, 2009). If I appear modest about our progress in the presence of such enormity, so I am. But let no-one presume that I am therefore not greatly enthused by the progress I believe I and my students (now colleagues) have made over these four decades. It has been a privilege to be allowed the time to work on such a most attractive and engaging topic.