Fitness for purpose: ac.vi.es Usability of data, Producer general - PDF document

FITNESS FOR PURPOSE LAND ECV RECORDS (D6.1, D6.2) QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017 Fitness for purpose: ac.vi.es • Usability of data, Producer general QA; general recommenda.ons • Demonstra.on of applicability for specific applica.ons (D6.2) • GCOS compliance (D6.1); temporal stability User QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017

General usability • Analysis of preliminary QA4ECV land surface datasets – Check of usability with standard tools like e.g. nco, cdo • CF compliance – Iden.fica.on of minor problems • Already feedback to data producers QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017 QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017

QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017 QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017

QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017 USING QA4ECV RECORDS FOR CMIP6 MODEL ASSESSMENT QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017

QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017 QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017

T6.2 Fitness for purpose Land (I) Model evaluaIon for CMIP6 CERES QA4ECV QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017 Differences QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017

Regional means (albedo) QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017 Examples for GlobAlbedo: Climatology difference QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017

Example: correlation albedo Kendalls-tau correla<on coefficient QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017 REGIONAL TO GLOBAL WATER AND ENERGY FLUXES QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017

Water and energy fluxes • Exploit high resolu.on and longterm datasets • Integra.on of QA4ECV products (albedo, faPAR) into HOLAPS model • Assessment against Fluxnet sites (local scale) • Water balance (catchment scale); runoff etc. • Use of uncertain.es to assess impact on uncertain.es of water fluxes QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017 Water and energy fluxes Loew et al., 2016 QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017

REGIONAL STUDIES QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017 What will be done? • Extreme events – Inves.gate fAPAR for drought detec.on over global scale and focus on the typical areas (China, Amazon, Russia) in 2010 • Longterm changes – Inves.gate the fAPAR trend over the last decades and find its drivers QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017

The global scale impacts of extreme heat wave in 2010 AghaKouchak et al., (2015) QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017 The excep.onal 2010 Amazon drought Xu et al., (2011) Lewis et al., (2011) Widespread decline in greenness of Amazonian vegeta.on due to the 2010 drought QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017

The 2010 spring drought in southwestern China Zhang et al., (2012) Reduced primary produc.vity Monthly air temperature and precipita.on averaged over southwestern China: (a) monthly total precipita.on (mm), (b) monthly mean temperature (°C). The 2010 spring drought can be tracked back to September 2009. The shaded area indicates the en.re dry period from September 2009 to May 2010 including the 2010 spring drought event. Loca.on of our study region within southwestern China. The land cover map is reclassified from the 500 m MODIS land cover map. Regional anomalies of (a) EVI and (b) GPP during spring (March– May) 2010 rela.ve to the means over the period 2000–2009. QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017 The 2010 Russian drought Yoshida et al., (2015) SIF and NDVI anomalies for 2010 Russian drought driven primarily by fPAR decline. QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017

a 3 a b 2 LAl anomalies (m 2 m − 2 ) 1 0 − 1 − 2 < − 15 − 10 − 5 0 3 6 9 12 15 18 >25 − 3 Trend in GIMMS LAl3g (10 − 2 m 2 m − 2 yr − 1 ) Year c QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017

Temporal stability? • ​" ( ​# − ​% ) /"' • DefiniIon (VIM 4.19, GCOS) a. abili.y of a data record to detect long-term trends b. property of a measuring instrument to provide similar measurements when the measurand remains constant in .me (dy/dt = 0) c. Can be expressed in absolute and rela.ve terms QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017 Trend detec.on, homogeniza.on tes.ng breakpoint QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017

Trend detec.on, homogeniza.on tes.ng • Time series analysis – Python package – Breakpoint detec.on (BP) – Par.al trend es.ma.on with BP – Homogeniza.on – Seasonality removal – Verbesselt et al. (2010) QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017 SAVS sites (>2500 sites) • Surface albedo valida.on sites (Loew et al., 2016) QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017

Example: GlobAlbedo QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017 Globalbedo trends Preliminary results QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017

Example: 5DTIP faPAR QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017 Conclusions • ESMValTool integra.on for CMIP6 ! Ini.al diagnos.cs for faPAR, albedo have been implemented – Further refinement foreseen once datasets available • GCOS assessment on temporal stability ! Tool has been developed and applied to pre-cursor datasets – Further refinement ongoing (snow effects, seasonality, correlated errors, minimum detectable trend …) • HOLAPS – WaiIng for mulIyear records • Regional studies have been defined – WaiIng for longterm records QA4ECV 2nd Technical (Interim) Review Mee<ng, EUMETSAT, Darmstadt, 21-22 February 2017