WatQ2E: Modeling Water Quantity, Quality and Ecosystems at optimal complexity
Platform Project - PP2.2
With WatQ2E, intelligent and comprehensible models are being developed, the results of which are to serve the protection of sufficient quantity and quality of water resources as well as the functionality of aquatic ecosystems. For this purpose, the models must be able to reliably represent the current and future behaviour of the resources on a management-relevant scale. This requires a mechanistic understanding of processes and patterns derived from observations. In this initiative, we jointly pursue the goal of using our models to find answers to scientifically but especially socially important questions under global change.
Background and scientific challenges
Water quantity and quality models are powerful tools for analysing the various hydrological, biogeochemical and ecological processes that determine water and matter fluxes in catchments and landscapes. In order to understand the behaviour of a resource in response to changing environmental conditions at small to supra-regional scales, the relevant processes, which mostly occur at small scales, must be understood and representable by our models at large scales as well.
There is usually a lack of reliable methods for parameterisation and regionalisation that enable the models to (i) allow reliable simulations on large scales and at different spatial resolutions, (ii) be transferable across different catchments, regions and climate zones, (iii) work even with low data availability and, most importantly for assessing the future, (iv) also be suitable for periods outside the calibration window. These are the biggest challenges we face.
In addition, we aim to improve the models to integrate the important interactions of resources with aquatic ecosystems and their function as fully as possible. Similarly, the models need to be able to simulate not only the evolution of traditional pollutants (e.g. nutrients) but also that of the omnipresent new pollutants.
The goal of improved parameterisation and regionalisation of the models will be achieved by optimising the models in combination with innovative methods of observation in groundwater as well as in rivers and lakes. This will enable us to reliably forecast the development of water quantity and quality as well as ecological impacts.
Only about 2.5% of the world's water is freshwater: one of the most important life-sustaining resources. But only about a fifth of it is available in the form of groundwater, rivers and lakes. Human activities and climate change are having increasingly serious impacts on the available quantity of water, its quality as well as on the associated ecosystems - at local, regional and even global level.
Against this background, we want to provide a reliable assessment of the current and future state of our water resources. To do this, we need to (further) develop the available modelling tools so that they map the important processes, integrate the mechanisms behind them only as complexly as necessary, and act quickly and accurately at the scales relevant to water resource management.
To properly understand and embed the key processes, dedicated process studies will be conducted using highly instrumented field sites and complex data sets (including isotopic and tracer data) from existing ( ) and developing ( ) observatories and monitoring platforms ( ).
Contents and envisioned outcome
These objectives and targets will be tackled in several sub-projects, oriented towards a variety of compartments and interfaces between them.
Development of a multiscale water Quality Model (mQM)
mQM is an integrated project aiming at providing a UFZ-wide platform for modelling of water quality variables at management-relevant scales. mQM leverages the hydrologic components of the existing mHM framework, while the modularized programming framework allows coupling to any other hydrologic model. The current work of mQM focuses on representing the dynamics of agro-chemical solutes (starting with N, followed by P and DOC) taking into account both biogeochemical and hydrologic legacies in soil, unsaturated zone, groundwater, and river reaches.
Processes are conceptualized with increasing complexity starting with a coarse model. For example, the transit time distribution (TTD) of groundwater is first represented by a lumped TTD at annual to monthly time scale, which later increase to daily resolution and eventually to a model with a spatially distributed groundwater representation. A first prototype model for nitrate using a TTD-based subsurface description has been developed at Selke River (Nguyen et al. 2020; WRR).
National GroundWater-Resource Model (GWRM)
The aims to reproduce and elucidate the current status and the future development of groundwater (GW) resources and their reliability as baseflow generator under climate and socioeconomic changes on a national scale. It starts with a system of regional GW-models, which eventually will be merged to assess national-scale groundwater resources being relevant for water economics and for farmland irrigation. GWRM is lined up to be used to assess and predict groundwater and surface water flow in terms of quantity, vulnerability and ecosystem services at management scale. This approach requires the development of innovative parameterization schemes based on combined examination of hydrochemical proxies, isotope-based residence time distributions and spectral analyses of GW-head and discharge time series.
The intended high spatial resolution of few hundreds of meters requires High Performance Computing (HPC) applying the open-source project . The existing framework will force the flow providing groundwater recharge and options to simulate future climate projections, which in turn will give insights into quantitative development and vulnerability of the resources considering prospective water demand of water works, farming and ecosystems.
This subproject advances the incorporation of isotope data into (i) hydrologic models to better constrain transit times and flow paths (e.g., stable water isotopes and absolute age tracers), and (ii) water quality and solute transport models to constrain turnover processes and rates as well as to fingerprint input sources (e.g., stable isotopes of N and O in nitrate). This subproject includes several activities:
- PhD-projects to derive transit times for events and water quality modelling across scales and contrasting catchments
- Utilization of stable isotope data in parsimonious models for nitrate age dating in an agricultural headwater catchment
- Bayesian stable isotope mixing models for assessing the dynamics of proportional contributions of different nitrate sources in mesoscale river catchments
- Tracer-aided ecohydrological and water flux modelling in the Bode catchment across scales using the EcH2Oiso model
- Nitrate isotope modelling to evaluate riparian denitrification
Optimum Complexity Modelling of Coupled Catchment-lake Biogeochemistry
Standing waters modify the flux of matter and the temperature of water flowing through the river network. Since most flowing waters are dammed or regulated in cultural landscapes, standing waters need to be parameterized in regional water quality models. In this subproject, we will couple a hydrodynamic-biogeochemical lake/reservoir model to a simple catchment model to characterize the retention of nutrients of standing waters under different event loading scenarios and lake/reservoir characteristics. The aim is to derive a simplified representation of the effect of dams on catchment biogeochemistry. The land-surface catchment contributions of water and solutes will be based on a data-driven modelling framework and the existing mHM and the envisioned mQM.
Modelling the Transport of Plastic from Land2River and River2Sea
Rivers are considered to be a major pathway for connecting land-based plastic sources with the oceans, while the magnitude of export of plastic debris is a combination of inputs and transport processes. This project tackles both separately: 1) Inputs (Land2River) and 2) Riverine transport (River2Sea). The first part focuses on providing a framework for synthesizing the information on point and diffuse sources for plastics of all sizes in rivers, while the second part deals with developing (and implementing) a modelling concept for transport of plastic debris in riverine systems.
We will leverage the graph (network) based modelling approach, which accounts for the micro plastic transport under steady state conditions, with further consideration of the retention and remobilization processes. These are crucial for the reliable depiction of plastic transport during extreme events like floods, which often mobilize a large fraction of the total loads within a short time span. The overall objective of this subproject is to quantify inputs of point and diffuse sources to better delineate time resolved hotspots where plastic debris enters the aquatic environment and is transported towards the oceans, which are the largest sink of plastic contamination. We will also better quantify transport in rivers both in terms of loads (relevant for global plastic flow assessment), concentration relevant for ecotoxicological assessment, and travel times (relevant for assessing legacy effects).
Event-based Model Calibration and Validation Approaches
Traditionally rainfall-runoff models are evaluated using performance metrics that measure the agreement between continuous observed and simulated streamflow time series (e.g., Nash-Sutcliffe efficiency), while ignoring that dominant generation processes may vary among different rainfall-runoff events. In cooperation with
River Aquatic Ecosystems Impact Modelling
Systematic identification of hot-spots for eutrophication in rivers from anthropogenic pressures is key to mitigate present and potential risks of impaired aquatic ecosystems. Comprehensive consideration for complex feedback of geomorphologic, hydrologic, biogeochemical, and anthropogenic conditions is necessary to simulate ecological dynamics of algal communities at the scope of the entire river basin. Especially for larger-scale or data-poor river basins, out-of-shelf models with huge equations of processes are not applicable to simulate algal dynamics due to limited data for calibration and verification. This stand-alone modelling project aims to improve a river-network-scale parsimonious model for the coupled complex algal-nutrient dynamics (CnANDY), facilitating the diagnosis of eutrophication and the provision of monitoring/management strategies with minimal amount of data required.