Presentation on theme: "CCI CMUG Integration Meeting 14.-16.03.2011 Rainer Hollmann Deutscher Wetterdienst Cloud cci."— Presentation transcript:
CCI CMUG Integration Meeting Rainer Hollmann Deutscher Wetterdienst Cloud cci
CCI CMUG Integration Meeting Overview of requirements Product specifications Uncertainties Consistency with needs of climate Research Group consistency with other ECVs (ECMWF data needs … are documented in DARD) Overview of talk
CCI CMUG Integration Meeting User requirements URD: compilation of requirements from the users point of view covering ideally the full spectrum of applications Product specification: The answer to the URD what is achievable and feasible and how it is designed
CCI CMUG Integration Meeting User Requirements Kent et al (2008): The definition of user requirements for the marine observing system is not a simple process. When asked, users have a tendency to ask "what can you provide?" or reply "as much as possible Cloud User requirements based on GCOS CMUG EPS-SG, MTG, CM SAF WMO DOR, Database of observational requirements ESA CCI Variables Globsnow (a few days ago)
CCI CMUG Integration Meeting Requirements as stated in GCOS-107, product A-4: GCOS requirements for clouds Requirements as stated in GCOS-107, product A-4, update 2011 Accuracy Spatial resolution Temporal resolution Stability (target) Justification Source Cloud Cover km3 hn/a Cloud height 500 m100 km3 hn/a Cloud temperature 0.3 K100 km3 hn/a Cloud ice profile n/a100 km3 hn/a Cloud water profile n/a100 km1-3 hn/a Accuracy Spatial resolution Temporal resolution Stability (target) Justification Source Cloud Cover km1-3 h0.01/decCFMIP Cloud pressure 50 hPa50 km1-3 h 0.25 K/dec NISTIR Cloud temperature 1 K50 km1-3 h 5 hPa/dec NISTIR Cloud ice profile 25 %50 km1-3 h2 %/dec NISTIR Cloud water profile 10 %50 km1-3 h5 %/decNISTIR- 7047
CCI CMUG Integration Meeting CCI projects Cloud variable Time-period Instruments / Satellites (e.g. Modis, AVHRR) Comments (any) Aerosols CCI Cloud cover2008, 1997A(ATSR)-2, MERIS, AVHRR/3, GOME(-2), SCIAMACHY, OMI, POLDER (1) the major goal are consistent cloud masks (avoiding double counting of pixels as cloud and as aerosol and allowing for a interim zone which is exploited by neither ECV) for ATSR instruments as baseline (2) More years proposed in options; long-term record feasible 1995 – 2020 Cloud Properties requirements from CCI variables (I)
CCI CMUG Integration Meeting CCI projects Cloud variable Time-period Instruments / Satellites (e.g. Modis, AVHRR) Comments (any) Sea Surface Temper ature CCI Cloud cover, cloud optical thickness (within Aug Dec 2010) AVHRR GAC and (A)ATSR full resolution A substantial sample only is required. Cloud cover to identify failures-to-detect in our cloud mask. Cloud optical thickness to assess if there is a critical thickness to be masked. Fire CCICloud cover, cloud top height ATSR, AATSR, VEGETATION, MERIS-RR, MERIS-FRS Cloud cover is needed for masking pixels. In addition, cloud top height would be of interest for cloud shadow identification. Cloud Properties requirements from CCI variables (II)
CCI CMUG Integration Meeting CCI projects Cloud variable Time-period Instruments / Satellites (e.g. Modis, AVHRR) Comments (any) Land Cover CCI x1: cloud cover, cloud top height x2: cloud optical thickness, cloud top height, cloud top pressure From 1998 (sure) From 1992 (to be confirmed) MERIS Full Resolution (from 2005) MERIS Reduced Resolution (from 2005) SPOT-VGT (from 1998) NOAA-AVHRR (from 1992 – to be confirmed) (A)ATSR (from to be confirmed) X1 - pixel identification X2 - atmospheric correction request: cloud shadow Cloud Properties requirements from CCI variables (III)
CCI CMUG Integration Meeting Cloud CCI URD is reflecting all important Req. has been forwarded to open review of science community (GEWEX RP) Needs of Cloud CCI Climate Research are included. Consistency with needs of Climate Research Group
CCI CMUG Integration Meeting Consistency with other CCIs Cloud CCI provided URD to all CCI SC for feeback received positive feedback from Aerosols, Land Cover, SST, Fire needs are documented in URD covering different time periods and instruments Most likely cloud CCI will not be able to cover all additional requirements Evaluation process ongoing
CCI CMUG Integration Meeting Consistency on Level 1 Radiances CCI URD addresses the GCOS need for FCDR for all instruments CCI Cloud is performing Level 1 Intercalibration for AVHRR-AATSR-MODIS instruments CCI Variables based on same instruments should agree on the same FCDRs!
CCI CMUG Integration Meeting Temporal resolution of final products (based on local satellite equator crossing time) AATSR/MERIS (ENVISAT): 10:00 am/pm AVHRR (N15, N16, N17, N18): am/pm (drifting), 02:00 am/pm (drifting), 10:00 am/pm, am/pm MODIS (AQUA,TERRA): 01:30 am/pm, 10:30 am/pm Providing the final products: Monthly means for each individual sensor Monthly means for each sensor group (e.g. all AVHRRs) Monthly means for all sensors on all satellites AATSR-MODIS-AVHRR AATSR-MERIS Current baseline: Specifications
CCI CMUG Integration Meeting Name Satellite data Temporal resolution MM at sat equator crossing time (L2 is swath based) Spatial resolution Accuracy Error estimates included in dataset GoalThreshold Cloud cover (high/mid/ low-level) AATSR AVHRR MODIS - AATSR (ENVISAT): equator crossing times - AVHRR (N15, N16, N17, N18): equator crossing times - MODIS (AQUA,TERRA): equator crossing times L3: 50 km L2: swath based MM: better than 0.08 cloud fraction or 10% bias 20% bc-rms MM: better than 0.15 cloud fraction or 20% bias 40% bc-rms Cloud top height AATSR AVHRR MODIS Same … L3: 50 km L2: swath based MM: 300m/300m/ 800m 1500m bc-rms MM: 800m/800m/ 1200m 3000m bc-rms Current baseline: Specifications (I)
CCI CMUG Integration Meeting Name Satellite data Temporal resolution MM at sat. equator crossing time (L2 is swath based) Spatial resolution Accuracy Error estimates included in dataset GoalThreshold Cloud top temperature AATSR AVHRR MODIS Same … L3: 50 km L2: swath based MM: 2/3/5 K bias 5/7/9 K bc-rms MM: 4/6/10 K bias 6/9/13 K bc- rms Liquid water path AATSR AVHRR MODIS Same … L3: 50 km L2: swath based better than 25 g/m2 or 15% bias 30% rmse better than 25 g/m2 or 30% bias 50% rmse Ice water path AATSR AVHRR MODIS SameL3: 50 km L2: swath based better than 25 g/m2 or 25% bias 40% rmse better than 25 g/m2 or 40% bias 70% rmse Current baseline: Specifications (II)
CCI CMUG Integration Meeting Product compliance (I) ID Horizontal Resolution (km) Temporal Resolution (h)Accuracy TSGCOSCMUGWCRPTSGCOSCMUG WCR P TSGCOSCMUG WCRP cloud CCI_cc G 100 BT 500 T 10 MD 30 TM 50 G 85.5 BT 250 T ten times daily % bias 20% bc- rms 10% 15% 20%5% 5% 7.9% 20% cloud CCI_cth G 200 BT 500 T 10 MD 30 TM 5 DA 100 G 171 BT 500 T ten times daily % bias 1500m bc-rms 0.5km 1.0km 2.0km 0.1km 0.2km 0.1km 0.5km 0.63km 1km cloud CCI_ctt G 200 BT 500 T 10 MD 30 TM 50 G 85.5 BT 250 T five times daily /3/5K bias 5/7/9K bc-rms 0.3K 0.4K 0.6K0.25K 0.5K 0.794K 2K
CCI CMUG Integration Meeting Product compliance (II) IDHorizontal Resolution (km)Temporal Resolution (h)Accuracy TSGCOSCMUGWCRPTSGCOSCMUG WCR P TSGCOSCMUG WCRP cloud CCI_lwp G 200 BT 500 T 10 MD 50 G 85.5 BT 250 T five times daily % bias 30% rmse 10(5)kg/m ² 20(10)kg/ m ² 50(20)kg/ m ² - 10kg/m ² 17.1kg/m ² 50kg/m ² cloud CCI_iwp G 200 BT 500 T 10 MD 50 G 85.5 BT 250 T five times daily % bias 40% rmse 10g/m ² 15g/m ² 20g/m ² -10g/m ² 12.6g/m ² 20g/m ²
CCI CMUG Integration Meeting Current baseline: Temporal resolution How to create a regular temporal sampling? Cloud CCI products will have level2b format global coverage products for each individual local solar time. combine all 14 orbits (per satellite & day) into two global data sets; (ascending /descending node having identical local solar times will give full global coverage in 6 hour time resolution (with two satellites differing 6 hours in local solar time). disadvantage in comparing with model data: post-processing to transform the information into the same representation.
CCI CMUG Integration Meeting Product compliance (III) Summary on compliance for Cloud CCI: Specification meet the GCOS 107 Requirements Further discussion on requirements from CMUG with TO & SC needed Joint work on cloud simulator approach Identifying case studies to supply full inst. data for process studies …
CCI CMUG Integration Meeting Focus on Level 2 uncertainties Reference data will be A-Train data and high quality surface networks Possibly include threshold limit to allow filtering of model data Started process of evaluation of Level 2 to Level 3 uncertainties supported by international science community (CREW, GEWEX Cloud Assessment) Aiming at radiative consistency on Level 3 Uncertainties
CCI CMUG Integration Meeting Error uncertainties Level 2 Non Systematic errors: Retrieval error: Uncertainty estimation of the retrieval system itself based on optimal estimation methodology. Sampling error: Due to the sampling of temporally and spatially unequally distributed samples. Systematic errors (bias) Radiometric Bias: calibration issue Retrieval Bias: Biases related to shortcomings in a retrieval itself. Sampling/Contextual Bias: Biases related to where a retrieval is/is not performed or contextually related uncertainty in a scene. Aggregation/Data Reduction Bias: Loss of required information during conversion to a gridded product or during analysis. Other considerations Cognitive Bias ? Correlated error "Independent" products that share similar biases; Tautology -Circular reasoning or treating non-independent data as independent during tuning.
CCI CMUG Integration Meeting
Temporal and spatial resolution N128 reduced Gaussian grid, ERA-Interim reanalysis at 00 UTC and 12 UTC ERA-Interim forecast 00 UTC +06h and +12h, ERA-Interim forecast 12 UTC +06h and +12h ParameterECMWF parameter descriptionAvailability at ECMWF (comments) Land/Sea maskLand-sea mask (172)As surface property Surface temperatureSkin temperature (235) Sea surface temperature (34) As surface property (skin temperature to be used for land surfaces; sea surface temperature to be used for sea surfaces) Sea ice maskSea-ice fraction (31)As surface property Surface pressureLog surface pressure (152)At model levels (requested for surface only) Snow albedoSnow albedo (32)As surface property 10m winds10 metre U wind comp. (165) 10 metre V wind comp. (166) As surface property Cloud CCI requirements from ECMWF (I)
CCI CMUG Integration Meeting Temporal and spatial resolution N128 reduced Gaussian grid, ERA-Interim reanalysis at 00 UTC and 12 UTC ERA-Interim forecast 00 UTC +06h and +12h, ERA-Interim forecast 12 UTC +06h and +12h ParameterECMWF parameter descriptionAvailability at ECMWF (comments) 2m temperature2 metre temperature (167)As surface property Temperature profileTemperature (130)At model levels (entire profile requested) Specific Humidity profile Specific humidity (133)At model levels (entire profile requested) Geopotential heightSurface geopotential (129)At model levels (requested for surface only; profile to be calculated) TCWV climatologyMonthly means of TCWV (137)As climatology Cloud CCI requirements from ECMWF (II)