Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Predicting lake responses to change using (near) real-time models Raoul-Marie Couture Donald Pierson (Leader) Eleanor Jennings Elvira de Eyto Erik Jeppesen Gideon Gal Vannforsk nettverkstreff og seminar Tirsdag 29. november, Clarion Hotel Royal Christiania, Oslo Consortium description ACRONYM JPI topic PROGNOS 2-extremes Predicting In-Lake Responses to Change using near real-time models INVESTIGATOR Coordination Partners Water quality monitoring; Sensor networks; Economically-optimized water management; DOC; Cyanobacteria blooms; Climate change. INSTITUTION COUNTRY Donald Pierson Uppsala University Sweden Raoul-Marie Couture and Isabel Seifert-Dähnn Norwegian Institute for Water Research – NIVA Eleanor Jennings Dundalk Insitute of Technology Ireland Elvira de Eyto Marine Insitute Ireland Erik Jeppesen and Dennis Trolle Aarhus University Denmark Karsten Bolding and Jorn Bruggeman Bolding & Bruggeman ApS Denmark Gideon Gal Israel Oceanographic and Limnological Research Norway Israel Stakeholders Stockholm Vatten (Water utlity, Lake Mälaren) Irish Water (Water Utility) Israeli Water Authority (Water utilty, Lake Kinneret) Bolding & Bruggman (SME) LakeLand Instruments (SME) Luode Consulting (SME) We are looking for a Norwegian stakeholder/case study Objectives Issues Relevant for Water Quality High DOC concentrations Algal Blooms Demonstrate the value of High Frequecy (HF) water quality monitoring to support water managment decisions Information on the present state of the lake Introduce HF data in water quality models and enable short-term water quality forecasts. Information on the near-future state of the lake Enhanced information to support water management decisions Leverage HF water quality monitoring data Why High Frequency Monitoring? Water quality is shaped by episodic events that are not easily captured by routine monitoiring programs These events occur along a continum of frequency and intensity Extreme events – large storms river flows etc Less extreme, yet important for affecting lake processes Wind gusts and stills Heating and cooling events Threshold Events – Stratification, Ice break-up Frequenct of episodic events expected to increase as a consequence of climate change Why Modelling? • HF monitoring reveals what is happening in real-time • Management must react fast • Most useful in system with directional water flow • What if modelling could predict near-future conditions, alike to weather forecast ? • Forecasts to react to upcoming e.g. thaw, rain event • Response (and resilience) to statistically likely extreme events • Model simulations can be improved if models are updated with HF monitoring data • Initial Conditions • Parameter values • Extract value from HF monitoring data by using with models Coupling of Monitoring and Modelling Weather Data Weather forecast Lake Water QualityModel Setup Simulation Short-Term Forecast Water Managment Decision Support Parameter Estimation Long-Term Forecast HF Monitoring Data Effect of Large Rain Event on Lake Feeagh, Ireland • • • • High rainfall June 21 Pulse of DOC to lake Deepening in stratification Decrease in stability that persisted for rest of summer Jennings et al. Freshwater Biology 57, 589-601, 2012 Effect of Wind Driven Upwelling Lake Erken Sweden Buoy Temperature Profiles HF monitoring data (Norwegian test site: Lake Langtjern) Flow, CDOM, Temp, pH On-site weather station Lake sensors: temperature and oxygen All-year lake buoy sensors at 8 depths Updated hourly at www.aquamonitor.no Lake water quality model - MyLake Saloranta and Andersen. Ecological Modelling, 2007, 207 (1), 45-60. Couture et al. Journal of Geophysical Research, 2015, 120 (11), 2441–2456. PROGNOS model - lake temperature vs time Planned infrastructure Innovation Forecasts for adaptive water management Methods of data assimilation into models Reducing model time steps to match HF data streams Routinely process data, run simulations, and produce forecasts Improve simulations of phytoplankton blooms Simulate DOC, with emphasis on drinking water quality Cost benefit analysis working with information from major water supplies Need for a Norwegian case study in PROGNOS Optimize water withdrawal and use We focus on lakes à DOC and algal bloom Added value of HF data stream Existing HF sensor arrays and data streams Towards a model-based decision support systems Cost Benefit Analysis Demonstrate the benefit of WQ forecasts for water treatment costs Network with peers in Sweden, Denmark, Ireland and Israel Join us in Oslo, June 13-15 2017 for the PROGNOS annual meeting NIVA’s team Acknowledgements Project management Raoul-Marie Couture (PL for Norway) Isabel Seifert-Dähn (WP5 leader) http://prognoswater.org/ @PROGNOSproject [email protected] Collaborators Jose-Luis Guerrero (hydrology, scientific computing) Heleen deWit (catchment DOC) Kari Austnes (exteme events) Funding