关键词: 邻苯二甲酸二辛酯, 环境暴露, 多介质, 时空迁移特征, 生态风险, SWAT模型

Abstract:

Global climate change and the emergence of persistent organic pollutants (POPs) pose growing challenges to ecosystem; how the regional risks by POPs will evolve within the context of climate change is a question of significant importance. This study presents a demonstrative investigation with di-n-octyl phthalate (DOP) in the Dianchi Lake Basin, a shallow urban lake with critical ecological functions, diverse pollution sources, and urgent restoration needs. Taking DOP as a model pollutant, we evaluated its ecological impacts under climate change, aiming to address key knowledge gaps in multimedia environmental modeling. To surmount the limitations of conventional watershed models, this study applies the SWAT-KM, a hydrologically based multimedia environmental model integrated with a shallow-lake module (the LKZ module). This model aims to characterize the spatiotemporal fate and transport of emerging contaminants, as well as their associated ecological risks, within coupled watershed-lake systems under changing climatic conditions. The study covers a historical period (2010-2024) and future scenarios (2025-2030) based on CMIP6 Shared Socioeconomic Pathways (SSPs). It aims to enhance the understanding of DOP’s multimedia exposure pathways and provide a rigorous, system-level foundation for urban-lake pollutant management. SWAT-KM (based on SWAT model, KM is a phonetic transcription of ‘chemical’), which was developed by the authors’ lab (MYB), was set up for the Dianchi Lake Basin integrated high-resolution geospatial datasets, land use and soil classification maps, and daily meteorological inputs which were obtained from ground-based observation stations and downscaled climate projections, while spatial layers were derived from authoritative geospatial databases and remote sensing products. SWAT-KM simulates pollutant transport and transformation across multiple environmental compartments on a daily basis with the spatial resolution of hydrologic response unit (HRU) level, as with SWAT model. Built upon the original SWAT hydrologic model, SWAT-KM creates an atmospheric module, a vegetation module, and harmonizes them with a suite of soil-and-aqueous processes, originally available in SWAT. After bearing the multimedia processes, both in individual phases and across the interphases, with chemicals’ adsorption/desorption kinetics, degradation kinetics, and transport kinetics where applicable, the SWAT-KM finally integrates the hydrological and atmospheric, with various chemical processes, into a dynamical multimedia model for certain chemical substance concentration estimation in each subbasin, and in the corresponding reach, air, and vegetation canopy, with daily updating. A key innovation of this study is the integration of the LKZ module into SWAT-KM to enable explicit simulation of shallow-lake dynamics with certain prescribed planar zone splitting. The lake was partitioned into hydrodynamic zones based on inflow locations, monitoring sites, and bathymetric features, which were dynamically linked to river segments within SWAT-KM model. Each zone receives daily inputs of runoff, sediment, and pollutant loads. Inter-zonal water exchange was calculated using the Chézy formula, and diffusion coefficients were estimated using a lacustrine heat budget method. The LKZ module solves a one-dimensional advection-diffusion equation to simulate the temporal evolution of pollutant concentrations, accounting for key source-sink processes, including atmospheric deposition, volatilization, photolysis/biodegradation, sedimentation, and benthic resuspension. The LKZ module addresses key limitations of traditional watershed models by providing a more realistic representation of in-lake hydrodynamics and pollutant transport, particularly at the water-sediment interface. Model calibration and validation relied on observed flow and sediment data, with performance evaluated using the coefficient of determination (R²) and Nash-Sutcliffe efficiency (NSE). Ecological risk was evaluated using the risk characterization ratio (RCR), defined as the ratio of the predicted environmental concentration (PEC) to the predicted no-effect concentration (PNEC) for DOP in the aquatic environment (8 ng·L⁻¹). Future simulations used CMIP6-based climate projections under four SSPs: SSP1-2.6 (low emissions), SSP2-4.5 (intermediate stabilization), SSP3-7.0 (regional rivalry), and SSP5-8.5 (fossil-fueled development). These scenarios were designed to disentangle the impacts of climate variability and extremes on DOP exposure dynamics. The results showed the following: 1) The SWAT-KM model effectively reproduced the spatiotemporal distribution of DOP. In river segments, the simulated mean concentration of dissolved DOP was 24.40 ng·L⁻¹, aligning well with observed values (9.35-54.50 ng·L⁻¹). In lake zones, the simulated mean concentration was 1.47 ng·L⁻¹, close to the measured average of 6.96 ng·L⁻¹, with short-term and localized deviations remaining within acceptable bounds. These results support the model’s suitability for simulating emerging contaminant exposure in urban lake basins. 2) DOP exhibited a distinct spatial pattern characterized by core-area accumulation and peripheral dilution, with high-risk zones concentrated near Kunming’s urban center. During the rainy season, dissolved DOP concentrations in rivers frequently exceeded 130 ng·L⁻¹, surpassing typical seasonal peaks (50-70 ng·L⁻¹), largely due to enhanced runoff and sediment resuspension. Persistent accumulation was observed in lake sediments, particularly in Zone 1 (the main inflow area), where concentrations approached 4.0 mg·kg⁻¹—substantially higher than in other zones (<0.90 mg·kg⁻¹)—suggesting a primary depositional sink. Multimedia simulations revealed strong inter-compartmental coupling across air, canopy, and soil systems, with seasonal dynamics driven by variations in precipitation, vegetation phenology, and atmospheric conditions. 3) The 2024 ecological risk assessment indicated that RCR_aqua values consistently exceeded unity across several reaches near Kunming urban center, suggesting sustained ecological pressure. Certain river reaches recorded RCR_aqua values above 5.0 during autumn and winter, highlighting localized high-risk hotspots. These elevated risks were primarily driven by cumulative urban non-point source pollution and domestic wastewater inputs, compounded by inadequate hydrodynamic dilution during the dry season. These findings underscore the need for targeted load reduction and focused management strategies in high-risk sub-basins, particularly during low-flow periods. 4) Under SSP5-8.5 and SSP3-7.0, more frequent and intense extreme precipitation events resulted in greater temporal variability in DOP concentrations. Peak dissolved concentrations in river and lake waters reached approximately 2.7 ng·L⁻¹, while sediment concentrations approached 5.0 mg·kg⁻¹. Enhanced resuspension during high-flow events markedly increased exposure risks. In contrast, simulations under SSP1-2.6 exhibited more stable exposure levels and reduced seasonal variability, underscoring the potential benefits of climate change mitigation in managing emerging contaminant risks. In conclusion, this study proposes a comprehensive multimedia simulation model that resolves cross-media interactions and in-lake processes, providing a robust technical basis for managing emerging contaminants in urban lake environments. Future lake-basin management should prioritize upstream source control, interception of contaminants-borne inflows, and sediment remediation, while also establishing early-warning systems to address risks from extreme climatic events and cross-compartment pollutant dynamics. Moreover, this module is readily transferable to other urban lake basins with similar pressures, offering a scalable, evidence-based approach for climate-resilient planning across diverse geographical and policy contexts.

Key words: Di-n-octyl phthalate (DOP), environmental exposure, multi-media, spatio-temporal migration characteristics, ecological risk, SWAT model