1 / 50

Andy Morse School of Environmental Sciences, University of Liverpool, Liverpool, U.K.

8.0 Applications 8.1 Status report on use of and need for research data in seasonal applications 8.1.1. Experience and progress from recent and ongoing projects (ENSEMBLES, UniCantabria Downscaling Portal, AMMA, QWECI ). Andy Morse School of Environmental Sciences,

ciel
Download Presentation

Andy Morse School of Environmental Sciences, University of Liverpool, Liverpool, U.K.

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. 8.0 Applications8.1 Status report on use of and need for research data in seasonal applications 8.1.1. Experience and progress from recent and ongoing projects (ENSEMBLES, UniCantabria Downscaling Portal, AMMA, QWECI) Andy Morse School of Environmental Sciences, University of Liverpool, Liverpool, U.K. A.P.Morse@liv.ac.uk CLIVAR WGSIP13 Buenos Aires, Argentina, 29-31 July 2010 Cyril Caminade, Dave MacLeod and Anne Jones, School of Environmental Sciences, University of Liverpool, Liverpool, U.K.; Matthew Baylis, School of Veterinary Science, University of Liverpool; Helene Guis, CIRAD, Montpellier, France.

  2. Background, Methods and Results, Discussion Introduction and Themes • Update and connects through research projects • Recent user experiences – NGOs and Government Research • Plots of distributed seamless activity (works in progress) • The climate services agenda

  3. Background, Methods and Results, Discussion Recent User Experiences • NGOs (major UK based international development and aid charities) • Humanitarian Futures Programme, Kings College London • http://www.humanitarianfutures.org/main/ • UK government bodies and commercial bodies in EQUIP and ENHanCE projects • African government programmes and decision makers through African partners • in QWeCI and HealthyFutures • How do we widen participation? • How do we leave climate information is a useable way through targeted narratives? • How does this experience link with current Climate Services Agenda initiatives?

  4. Introduction, Methods and Results, Discussion Seasonal Scales

  5. Introduction, Methods and Results, Discussion Seasonal Ensemble Prediction

  6. Introduction, Methods and Results, Discussion Potential Seasonal Skill in Epidemic Zones for Malaria Based on the Liverpool Malaria Model simulations driven by seasonal ensemble multi-model outputs (Rainfall and Temperature) ENSEMBLES Seasonal EPS May 4-6 (ASO) upper tercile epidemic transmission zone ROCSS

  7. Introduction, Methods and Results, Discussion Seasonal prediction of malaria epidemic risk in West Africa Potential skill using ENSEMBLES re-forecasts to drive a malaria model Skill of multi-model forecasts derived from ENSEMBLES May start date, averaged over 15 high-variability incidence grid points in WA. (15N: 17.5W to 7.5W, 12.5N: 2.5W to 12.5E, and 5N: 10E to12.5E). Standard error in brackets. Measured relative to skill of NCEP reanalysis-driven simulations.

  8. Introduction, Methods and Results, Discussion Seasonal prediction of malaria epidemic risk in West Africa Potential skill using ENSEMBLES re-forecasts to drive a malaria model ROCSS Model Skill of above the median forecasts for LMM-simulated incidence over May forecast months 4-6, 13 high-variability grid points in WA. (15N: 17.5W to 7.5W, 12.5N: 2.5W to 12.5E, and 5N: 10E to12.5E). Measured relative to skill of NCEP reanalysis-driven simulations . Scatter points show grid point values, solid black circles show areal mean.

  9. Introduction, Methods and Results, Discussion Dabbling with Decadal are working on in and

  10. Temperature [K] Distribution of ensemble members from first 5 years of ENSEMBLES decadal forecasts, observations: NCEP reanalysis Correlation coefficient (ensemble mean vs obs) = 0.233

  11. Predicting the AMO? Using the first 5 years of decadal hindcast experiments (except for the final 2005-2015 forecast) after van Oldenborgh et al 2010 GRL

  12. Climate Diagnostics Working with users on recent climate variability and trends, towards producing climate products, filling decadal gap with RCM runs Start with recent past climate using high resolution Eobs for Europe, ENSEMBLES RCM runs for Europe Need capture real variability – do these runs SRES GCM-RCM runs have even average variability? Should we even ask that question of GCM-RCM runs? How do we use RCM to fill decadal gap? RCMs with s2d initial condition ensembles?

  13. Introduction, Method, Climatic Trends, Health Impact examples Observed Climatic Trends: 1961-2004 Wetter and warmer winters over Northern Europe, warmer and drier winters over Southern Europe. More drought conditions over the Mediterranean basin in summer

  14. Introduction, Method, Climatic Trends, Health Impact examples Future Changes: 2030-2050 vs 1960-2000 Warming, faster over northern Europe in winter and southern Europe in summer. The winters get wetter over northern Europe for both seasons. Strong drying signal over the Mediterranean basin in summer. Shading: changes Dots: 80% of the climate models agree on the sign of changes

  15. Introduction, Method, Climatic Trends, Health Impact examples Recent climate T2m PDF JJA 1961-2000 EOBS observation in black ENSEMBLES RCM CTL ensemble (ERA40 driven) in blue ENSEMBLES RCM SRESA1B ensemble (GCM driven) The envelope d(red thin lines) depicts the spread (2stddev) of the CTL (SRESA1B) model ensemble with respect to the mean

  16. Introduction, Method, Climatic Trends, Health Impact examples T2m PDF JJA 2030-2050 vs 1961-2000 ENSEMBLES RCM SRESA1B ensemble (GCM driven) 1961-2000: Orange 2030-2050: Red The envelope (thin red lines) depicts the spread (2stddev) of the model ensemble with respect to the mean -> shift to warmer summers -> spread increases in the future

  17. Health Impact examples for Europe

  18. Health Impact examples: Bluetongue over Europe Mean Bluetongue Risk (OBS): ASO 1961-2008 High BT risk over Spain, Portugal, south western France, Sardegna and Sicilia. This misses out observed outbreaks in Corsica Unrealistic values over mountains and Eastern Europe Shading: Ro risk (arbitrary scaled between 0 and 1) From Guis et al, 2010

  19. Health Impact examples: Bluetongue over Europe Bluetongue Risk changes: 2030-2050 vs 1961-2000 MULTI-MODEL SPREAD: MAGNITUDE MULTI-MODEL SPREAD: SIGN CONSISTENCY MULTI-MODEL CHANGES: MEAN The BT risk increases over UK, Southern France and North-western Spain (Galicia) Changes in Northern Europe are related to the pathogen properties Changes in Southern Europe are associated with the spread of the Afro-Tropical vector (Imicolaspp) From Guis et al, 2010

  20. Health Impact examples for Africa

  21. Health Impact examples: Malaria Climatic Risk over Africa Mean seasonal cycle 1990-2007 Hovmoeller like diagram (zonal average between 16W and 16E) Shading: Rainfall Contours: Malaria Incidence Underestimation of the northern extension of the malaria incidence belt by LMM 2-3 months LAG between rainfall and malaria incidence

  22. Health Impact examples: Malaria Climatic Risk over Africa Mean annual malaria incidence 1990-2007 Endemic areas >80% “Endemic and seasonal” areas between 20-80% Epidemic Areas (<20%) -> Northen fringe of the Sahel -> Strongly connected to climate variability Underestimation of the Northern extension of the malaria incidence belt by LMM ITCZ extends too far north in the RCM world Mean annual simulated Malaria Incidence (1990-2007) driven by “Observed datasets” and the ENSEMBLES RCM ensemble

  23. Health Impact examples: Malaria Climatic Risk over Africa Mean Incidence changes SON 2031-50 vs 1990-2010 Simulated changes in Malaria Incidence (SON) based on the different RCMs -> common feature: decrease of the Malaria Incidence at the Northern fringe of the Sahel -> Related to changes in the number of rainy days (and not the seasonal amounts)

  24. Health Impact examples: Rift Valley Fever over West Africa RVF climatic risk 1990-2007 RVF risk RVF risk Ndione et al, 2008 Caminade et al, 2010 (in review) • Dry spell followed by a rainfall peak • during the late rainy season (Sep-Oct) • over Northern Senegal • Rehydrating ponds • mosquitoes hatching + hosts availability •  high RVF risk Rift Valley Fever risk (%) based on rainfall from ERAINTERIM reanalysis (1990-2007). The number of RVF risk events is defined by a dry spell (10 consecutive days with total rainfall below 1mm) followed by a convective event (high precipitation defined by one or two days following the dry spell above the 90th percentile) occurring during the late rainy season (SON).The total number of RVF risk events is then rescaled to range between 0 and 100% to define the risk. The dotted, crossed and filled black areas depict animal host densities (cattle + buffalo + sheep + goats) above 1, 10 and 100 per km2 (FAO, 2005).

  25. Health Impact examples: Rift Valley Fever over West Africa Synoptic situation: Senegal RVF outbreak 2002 OLR Hovmoeller Diagram (averaged between 12°N and 18°N). OLR Anomaly for 2002 (NCEP). Brown: Convective event Black Box: Senegal location 2 weeks predictability??? -> Value of medium range forecasts RVF outbreak 10-15 days predictability?

  26. Introduction, Methods and Results, Discussion Climate Services Agenda - a seamless one? • Who is doing what, where and with whom e.g. can we share good practise and/or join forces? • Are Met Services interacting as much as possible with other researchers working on impacts and data use? • Who is thinking seamlessly across multiple timescales? • Who is developing this seamlessness with the user community? • How can WGSIP, CLIVAR, WCRP help connect this Agenda with the impacts community?

  27. Introduction, Methods and Results, Discussion Summary to seamlessness • Grand ensemble approach – combined ensembles from different systems – bound uncertainty, maximise skill, model climates • Impacts model portability – develop models work different climate streams and grand ensemble – impact uncertainty, integrated model value • Field and Environmental Observations – verification and dynamic insight • Model data post processing – downscaling, bias correction, dressing • Continuity to society – decision makers, product tailoring, decision support systems, understanding critical uncertainty and thresholds, agent based models, combination with remote sensing and observations, through to policy and project impact on society

  28. QWeCI FP7 SEVENTH FRAMEWORK PROGRAMMETHEME ENV.2009.1.2.1.2Methods to quantify the impacts of climate and weather on health in developing low income countries Collaborative Project (small- or medium scale focused research project) for specific cooperation actions (SICA) dedicated to international cooperation partner countries Quantifying Weather and Climate Conditions on health in developing countries (QWeCI) 3.5 MEu EC contribution (~4.7MEu total) 1st Feb 2010 start 13 partners = 7 Africa, 6 EU, Liverpool coordinator, 42 months UNILIV, CSE, CSIC, ECMWF, IC3, ICTP, ILRI, IPD, KNUST, UCAD, UNIMA, UOC, UP

  29. http://www.equip.leeds.ac.uk/

  30. Improving the use of climate prediction by quantifying, understanding and managing uncertainty. Through working with stakeholders, the EQUIP team will develop new methodologies and analyses for using climate information that will be employed by decision makers in a set of case studies. We will quantify and understand the uncertainty surrounding future droughts, heatwaves, crop production and marine ecosystems. Introduction

  31. EQUIP: End-to-end Quantification of Uncertainty for Impacts Prediction • Edinburgh, Newcastle, Liverpool • NERC directed research • EQUIP network (external users and academics) is a core part of our research

  32. Introduction, Methods and Results, Discussion Extra Slides

  33. Background, Methods and Results, Discussion Integrated Climate Model Impacts Verification Paradigm from Morse et al. (2005) Tellus A 57 (3) 464-475

  34. Introduction, Methods and Results, Discussion Relative contributions of uncertainties Internal variability Scenario uncertianty Model uncertainty Hawkins & Sutton, 2009, BAMS, 90(8):1095-1107

  35. Introduction, Method, Climatic Trends, Health Impact examples UK Rainfall and Temperature Trends: EOBS Rainfall 1961-2004 Temperature 1961-2004

  36. Introduction, Method, Climatic Trends, Health Impact examples UK extremes in winter: EOBS Frost days: 1961-2004 Heavy rainy days: 1961-2004

  37. Introduction, Method, Climatic Trends, Health Impact examples Recent climate T2m PDF DJF 1961-2000 EOBS observation in black ENSEMBLES RCM CTL ensemble (ERA40 driven) in blue ENSEMBLES RCM SRESA1B ensemble (GCM driven) The envelope d(red thin lines) depicts the spread (2stddev) of the CTL (SRESA1B) model ensemble with respect to the mean Problems with 2 models (freezing days too frequent)

  38. Introduction, Method, Climatic Trends, Health Impact examples T2m PDF DJF 2030-2050 vs 1961-2000 ENSEMBLES RCM SRESA1B ensemble (GCM driven) 1961-2000: Orange 2030-2050: Red The envelope (thin red lines) depicts the spread (2stddev) of the model ensemble with respect to the mean -> shift to warmer winters -> spread increases in the future

  39. Introduction, Method, Climatic Trends, Health Impact examples Observed Climatic Trends: extremes 1961-2004 Largest Increase of winter rainfall extremes over the western coasts of the UK and Norway. Decrease in the number of frost days in winter over Europe Significant increase of warm days and warm nights over the Mediterranean basin in summer  Health Impacts

  40. Introduction, Method, Climatic Trends, Health Impact examples Observed Climatic Trends extremes Wetter and warmer winters over Northern Europe, warmer and drier winters over Southern Europe More drought conditions over the Mediterranean basin in summer

  41. Health Impact examples: Bluetongue over Europe Bluetongue Risk ASO, Northern Europe Simulated Relative Ro BT changes (with respect to 1961-2000) over Northern Europe. Black: BT risk based on EOBS Blue: BT risk based on CTL exp Red: BT risk based on SRESA1B exp The relative envelope depicts the spread within the RCMs ensemble (1 Stddev) 2006: BT outbreak in France Benelux and Germany captured by EOBS and the CTL exp Increasing trend for the future over Northern Europe From Guis et al, 2010

  42. Health Impact examples: Liver Fluke over the UK Liver Fluke: Fo=f(T,Rdays) 106 Fo: The total predicted number of adult progeny arising from pasture contamination by a single fluke present in a non-immune sheep for one year. X axis: Temperature (°C) Y axis: Fraction of rainy days (1 means it rains every day, 0 no rain, based on the 1mm threshold). 103 102 Work with J. Van Dijk

  43. Health Impact examples: Malaria Climatic Risk over Europe Mean Malaria “climatic” Risk: JAS 1990-2008 Based on LMM simulations driven by observed Rainfall and Temperature from different observed datasets. Northern Italy, some parts of Galicia in Spain and the “Landes” region in France are climatically “at risk” The incidence values are relatively low in magnitude (20-50%) compared to what can be expected in Africa.

  44. Health Impact examples: Malaria Climatic Risk over Africa Shift of the epidemic belt 2031-50 vs 1990-2010 Gray: Location of the epidemic belt 1990-2010 Black dots: Future location of the epidemic belt 2030-2050 The epidemic belt location is defined by the coefficient of variation, namely: Mean Incidence > 1% 1stddev > 50% of the average Southward shift of the epidemic belt over WA -> to more populated areas...

  45. Health Impact examples: Rift Valley Fever over West Africa RVF climatic risk based on RCMs: 1990-2007 RVF Risk based on the control RCMs ensemble (runs driven by ERAINTERIM at the boundaries). The analysis is carried out for the period 1990-2007. Problems as most of the models overestimate the northward extension of the ITCZ. KNMI and DMI pattern relatively realistic with respect to the reanalysis / GPCP driven runs. RVF risk estimated from ERAINT

  46. Health Impact examples: Rift Valley Fever over Africa RVF distribution map according to the Central for Disease Control and prevention. Blue areas show where RVF is endemic RVF risk as estimated from a) ERAINTERIM and b) GPCP for the whole African continent.

  47. Introduction, Methods and Results, Discussion Title Here

  48. Introduction, Methods and Results, Discussion Title Here

  49. Introduction, Methods and Results, Discussion Title Here

  50. Introduction, Methods and Results, Discussion Title • Disease

More Related