globalbedo.org

Executive summary of progress: 08/07/10 - 08/08/10
Executive summary of progress: 08/06/10 - 05/07/10
Executive summary of progress: 05/05/10 - 05/06/10
Executive summary of progress: 05/03/10 - 05/05/10
Executive summary of progress: 05/02/10 - 05/03/10
Executive summary of progress: 05/01/10 - 05/02/10
Executive summary of progress: 05/11/09 - 05/12/09

Executive summary of progress: 22/11/12 - 04/03/13

Since the last report, albedo products for 2001 - 2003 have been completed and 2000 is mostly completed. There has been some speed-up of the processing due to the acquisition (paid for from a separate EU-FP7 source) of 4 additional blades at the end of January 2013. Unfortunately, this has been somewhat offset by upgrading the operating system to RHELv6.2 (from v5.6) which has slowed down the processing due to memory issues with the Java Virtual machine. These have not yet been resolved. BRDF Products for 2005 - 2009 have been staged at CEMS although progress has halted recently due to the lack of effort available. MODIS MCD43A1,2 BRDF 500m products have not yet been transferred to CEMS for processing but the Python re-write of the original IDL code has been completed, albeit not yet tested at CEMS. UCL is still waiting for the AODs from Swansea University before they can start validating all the SURFRAD & FLUXNET sites for 2004 - 2012, where available. ESA granted the proposed contract extension to explore the retrieval of sea-ice albedo from AATSR/MERIS and compare these to MISR albedo retrievals in a modification planned by JPL. Feedback for the EU-FP7-QA4ECV project proposal to extend albedo to 30 years using GEO (amongst several other ECVs) was received on 28/2/13. This feedback was positive but only garnered 12/5/15, above the 10/15 threshold but not a guaranteed “shoe-in”. It is unclear what competition the bid is up against so it is currently unknown if contract negotiations will commence. Two abstracts were submitted on GlobAlbedo (main project and CCN on sea-ice) for the ESA “Living Earth” symposium to be held in Edinburgh in September 2013.

Executive summary of progress: 05/10/12 - 21/11/12

As previously reported, GlobAlbedo processing continues to be much slower than originally anticipated (a knock-on effect of previously mentioned Blue ice delays and problems with the commissioning of the supercomputer system at MSSL). 2004 & 2006 have been completed since the last report and all the products from 2005 - 2011 are now online. 2004 browse data and animations are currently being processed and it is expected that 2006 will be online by 30.11.12 at the latest. This means that by the end of November 2012 that 7.5 years will be complete (1/2004 - 6/2011).

MSSL submitted a proposal to the ISIC CEMS (Facility for Climate and environmental Monitoring - Harwell) in order to try and secure some free super computer time to assist with the processing of the GlobAlbedo data sets. Unfortunately, the original proposal was not accepted as the time pressure and the amount required by ISIC was considered beyond the resources available. However, with support from NCEO and the UK-TSB, parts of this proposal have just been accepted. The parts include (a) staging all the intermediate BRDF INVERSION and MERGED products on the ISIC data-store as there is no room for these at MSSL; (b) making these accessible through the GlobAlbedo website using wget commands that would allow users to download these data; (c) making the MODIS BRDF Prior available including processing all 500m land spectral bands.

We continue to receive valuable feedback on product quality from our users with MPI Hamburg providing the best inputs. We have re-processed a few datasets to remove artefacts. We plan to re-process 2010 - 2011 once we have completed the first cycle back to 1998. We are working towards completion of 2002-2003 by the end of the year with the remaining 5 years processed by March 2013. Due to external funding, our linux cluster will be enhanced by 4 additional blades in January/February 2013 so we hope to speed up and complete the first cycle by the end of March 2013. We have recently added Geoland2 data for 2006-7 to our validation plans but await the delivery of other years before the end of the year. UCL is still waiting for the AODs from Swansea University before they can start validating all the SURFRAD & FLUXNET sites for 2004-2012, where available.

Further SPOT-VGT data has arrived up until the end of 2011 and we hope to receive Jan-Jun 2012 early in the New Year. We hope to persuade SU to process these data and FUB to process MERIS data from July 11 - April 12 so we can extend the GA time series to June 12.

A proposal for enhancements of the GA products was submitted to ESA on 29.10.12. A proposal for a continuation of the GA approach but to 30 years of GEO+LEO was submitted to the EU-FP7 programme on 21.11.12. This took so much time that it was not possible to write a paper for the ESA Sentinel-3 workshop.

Executive summary of progress: 07/09/12 - 05/10/12

As previously reported, GlobAlbedo processing continues to be much slower than originally anticipated (a knock-on effect of previously mentioned Blue ice delays and problems with the commissioning of the supercomputer system at MSSL). 2007 is completed and it is expected that 2006 will be complete by mid October so that 6.5 years would then be complete (1/2005 - 6/2011).

MSSL has submitted a proposal to the ISIC CEMS (Facility for Climate and Environmental Monitoring - Harwell) in order to try and secure some free super computer time to assist with the processing of the GlobAlbedo data sets. The plan would be to port the processing environment at MSSL to a virtual cloud at ISIC-CEMS (on 40% of the MSSL's current system) and have it operating in parallel so that all the remaining years would be completed by the end of 2012. The UK Met Office is the latest user to formally declare they are evaluating the GlobAlbedo products.

Executive summary of progress: 03/08/12 - 07/09/12

As previously reported, GlobAlbedo processing continues to be much slower than originally anticipated (a knock-on effect of previously mentioned Blue ice delays. MSSL has approached CEMS (Harwell) in order to try and secure some super computer time to assist with the processing of the GlobAlbedo data sets. MSSL plans to issue a proposal to ICIS) by mid-September - free use award. Tapes are currently being prepared (2007 - 2011) for sending to ESRIN. 2007 data processing completed.

MSSL submitted a proposal to the ISIC CEMS (Facility for Climate and Environmental Monitoring - Harwell) in order to try and secure some free super computer time to assist with the processing of the GlobAlbedo data sets. Unfortunately, the original proposal was not accepted as the time pressure and the amount required by ISIC was considered beyond the resources available. However, with support from NCEO and the UK-TSB, parts of this proposal have just been accepted. The parts include (a) staging all the intermediate BRDF INVERSION and MERGED products on the ISIC data-store as there is no room for these at MSSL; (b) making these accessible through the GlobAlbedo website using wget commands that would allow users to download these data; (c) making the MODIS BRDF Prior available including processing all 500m land spectral bands.

We continue to receive valuable feedback on product quality from our users with MPI Hamburg providing the best inputs. We have re-processed a few datasets to remove artefacts. We plan to re-process 2010 - 2011 once we have completed the first cycle back to 1998. We are working towards completion of 2002 - 2003 by the end of the year with the remaining 5 years processed by March 2013. Due to external funding, our linux cluster will be enhanced by 4 additional blades in January / February 2013 so we hope to speed up and complete the first cycle by the end of March 2013. We have recently added Geoland2 data for 2006 - 7 to our validation plans but await the delivery of other years before the end of the year. UCL is still waiting for the AODs from Swansea University before they can start validating all the SURFRAD & FLUXNET sites for 2004 - 2012, where available. Further SPOT-VGT data has arrived up until the end of 2011 and we hope to receive Jan - Jun 2012 early in the New Year. We hope to persuade SU to process these data and FUB to process MERIS data from July 11 - April 12 so we can extend the GA time series to June 12.

A proposal for enhancements of the GA products was submitted to ESA on 29.10.12. A proposal for a continuation of the GA approach but to 30 years of GEO+LEO was submitted to the EU-FP7 programme on 21.11.12. This took so much time that it was not possible to write a paper for the ESA Sentinel-3 workshop.

Executive summary of progress: 07/06/12 - 03/08/12

As previously reported, GlobAlbedo processing continues to be much slower than originally anticipated (a knock-on effect of previously mentioned Blue ice delays. 2008 - 2011 is completed and available online and 2007 processing will start week beginning 5/8/12.

Executive summary of progress: 07/05/12 - 07/06/12

As previously reported, GlobAlbedo processing continues to be much slower than originally anticipated (a knock-on effect of previously mentioned Blue ice delays). The current processing status is summarised in Table 1.

	MODIS PRIOR	VGT BBDR	MERIS BBDR	ALBEDO	VALIDATION
2011		5	5	5	3
2010	5	5	5	5	3
2009	5	5	5	5	2
2008	5	5	5
2007	5	5	5
2006	5	5	5
2005	5	5	5	5	5
2004	5	5	5
2003	5	5	5
2002	5	5	5
2001	5	5
2000	5	5
1999		5
1998		5

Table 1: GlobAlbedo processing status (5-complete)

Executive summary of progress: 19/04/12 - 17/05/12

The updated GlobAlbedo web site has now gone live - feedback if generally very positive. A few minor background updates are still being completed.

As previously reported, GlobAlbedo processing continues to be much slower than originally anticipated (a knock-on effect of previously mentioned Blue ice delays). The current processing status is summarised in Table 2.

Concerning processing: currently halfway through re-processing 2009 and the start of 2008 processing is imminent.

All VGT & MERIS BBDR is now downloaded and available at MSSL

	MODIS PRIOR	VGT BBDR	MERIS BBDR	ALBEDO	VALIDATION
2011		5	5	5	2
2010	5	5	5	5	2
2009	5	5	5	4
2008	5	5	5
2007	5	5	5
2006	5	5	5
2005	5	5	5	5	5
2004	5	5	5
2003	5	5	5
2002	5	5	5
2001	5	5
2000	5	5
1999		5
1998		5

Table 2: GlobAlbedo processing status (5-complete)

Executive summary of progress: 01/01/12 - 19/04/12

As previously reported, GlobAlbedo processing continues to be much slower than originally anticipated (a knock-on effect of previously mentioned Blue ice delays). The current processing status is summarised in Table 2. It is fully anticipated that GlobAlbedo processing should now return to nominal speed following the return of the Blue ice processing system to nominal operations.

The project continues to suffer from issues associated with the MODIS priors, primarily at high latitudes concerned with the presence or absence of snow; this is especially the case near the South Pole where workarounds failed to eliminate the problem completely. The re-processing of the MODIS PRIORS was completed over the Christmas holiday and into the early part of January. At the end of January, it was decided that the test product was the best possible that could be achieved without sinking substantial additional effort into this and subsequently, the final albedo validation was completed of the global products in February. AOD validation for 2005 was pushed-back to enable BlueIce processing to be completed for BBDR (see Table 2). After two months, all the VGT and MERIS BBDRs have been downloaded onto the RAID store at MSSL ready for the global processing. AOD validation was completed by mid-March and the final version of the test product validation was completed and delivered shortly afterwards. This has now closed-out all outstanding items from the QR (Table 7-2).

The time variation of the MODIS-VGT-MERIS inter-sensor calibration carried out by the CNES SADE team was provided to the PI. The PUM was updated to take account of the actual situation in terms of output global mosaicked products only being available in lat,lon and tile products in MODIS SIN projection. Significant work has been done on the project website to make it fit for purpose to display (via still and animated images) all of the products and allow a qualitative assessment of product quality. In addition, a wget based system has been set-up to allow users to extract any of the completed products for any time period required within a year or for all years on either 8-day time-steps or monthly. Work is almost complete on the single point through time and a small area around a point for those users interested only in a small area.

The current processing progress means that one processing year takes approximately 18 days for production and 3 days for the generation of all the quick-look products and mosaics as well as assessment of the quality. After having detected an issue with the processing software for the end of 2010 (incorrect snow cover), it appears that the problem is back again for the end of 2009 in northern polar autumn. This will mean re-processing of these data in parallel to 2008. METEOSAT First generation data has been downloaded for inter-comparison with GlobAlbedo and MISR level-2 data is in the process of being ordered and downloaded for the remaining years of 2000 - 2011. A further batch of MODIS data products will be collected from NASA GSFC during the MODIS team meeting (7 - 9 May 2012). With the current processing capabilities, the remaining 10 years of products will take a further seven months to process. Steps are being taken at present to determine whether the computing facilities at UKSA-ISIC/ESA Harwell may be able to be used as BlueIce has been de-commissioned and the FUB computing facilities are fully committed.

With this current schedule slip, the UCM-2 will have to be delayed until October 2012 (to be confirmed with ESRIN); this is to give sufficient time for the validation to be completed and for the final products to be provided to users for their initial evaluation. Venue is TBD (but possibly ECMWF or the UK Met Office; to be discussed with ESRIN).

User requested data has been sent to ITC (Test Product all tiles at 1km in BEAM format), U. of Thessalonki (ASCII BRDF coefficients for specified test site), Météo-France (Test Product all tiles at 1km in BEAM format) as well as Thomas Kaminski and Bernard Pinty (Test Product all tiles at 1km in BEAM format). A full set of test products was sent out to ITC (Bob Su) and MPI Hamburg (Alexander Loeuw) on user-supplied disk in BEAM format as well as to Simon Pinnock on LT04 tape in the final netCDF format.

Executive summary of progress: 23/11/11 - 31/12/11

As previously reported, GlobAlbedo processing has been somewhat slower than originally anticipated (partly as a result of Blue ice delays). The current processing status is summarised in Table 3. It is fully anticipated that GlobAlbedo processing should now return to nominal speed following the return of the Blue ice processing system to nominal operations.

	MODIS PRIOR	VGT BBDR	MERIS BBDR	ALBEDO	VALIDATION
2011		5	5
2010	5	5	5
2009	5	3	5
2008	5		5
2007	5		5
2006	5		5
2005	5		5	4	4
2004	5		5
2003	5		5
2002	5		5
2001	5
2000	5
1999
1998
1997

Table 3: GlobAlbedo processing status (months)
The project continues to suffer from issues associated with the MODIS priors, mainly at high latitudes concerned with the presence or absence of snow; however, workarounds are in the process of being performed after review. It is planned that the re-processing of the MODIS PRIORS should be completed over the Christmas holiday. As soon as the test product is completed, the final albedo validation will be completed of the global products. AOD validation for 2005 was pushed-back to enable BlueIce processing to be done for BBDR (see Table 3). This is faster than previously due to the use of the orbit-based BEAM GPF processor. AOD validation is due to be completed by 20/1/12 so the final version of the test product validation can be completed and delivered shortly afterwards. This will then close-out all outstanding items from the QR. All MERIS BBDR data has been delivered to UCL from FUB. VGT BBDR will be ftp-pushed from SU once disk partitions are set-up in the New Year.

The GlobAlbedo consortium requests that it is permitted to apply 2008/09 Radcal inter-sensor coefficients used for narrowband-to-broadband conversion to all years (1998 - 2011). This is required as a result of CNES not providing SADE coefficients (MERIS to VEGETATION for 1998 - 2007 and 2009 - 2011).

ESA (ESRIN) has agreed that the RID is closed for the PUM.

It is still proposed that UCM-2 will have to be delayed as a result of the various processing delays to 30 April / 1 May 2012 (to be confirmed with ESRIN); this is to give sufficient time for the validation to be completed and for the final products to be provided to users for their initial evaluation. Venue is TBD (but possibly ECMWF; to be discussed with ESRIN).

User requested data has been sent to ITC (Test Product all tiles at 1km in BEAM format), U. of Thessalonki (ASCII BRDF coefficients for specified test site), MPI Hamburg (Test Product Europe only tiles at 1km in BEAM format) Météo-France (Test Product all tiles at 1km in BEAM format) as well as Thomas Kaminski and Bernard Pinty (Test Product all tiles at 1km in BEAM format). However, a new set of products will be sent out on user-supplied media once the re-processing of the 85 tiles is completed.

Executive summary of progress: 30/06/11 - 30/09/11

Progress has continued in terms of preparation and almost final completion of the QR, which began with a 120 hour test run on 22/7/11. Unfortunately the 120 hour test failed due to some issues with the software (system parameters). A repeat was held starting on the 22nd August 11 and completed successfully in time for the QR meeting held on 6th September 11 by video conference. It was worthwhile to note that the test performed (both times) twice as fast as originally planned, completing in just 60 hours.

All the deliverable documents were submitted for the QR (section 7.2). Some of these were accepted in a delta-QR held via video-conference on 26th September 2011.

Following formal agreement from ESRIN, the GlobAlbedo consortium started the processing of final data products. this is partly in order to meet the schedule constraints of PM5 (due late 2011).

Significant delays (3 weeks on top of 2 weeks processing), due to RAID I/O at MSSL were experienced which appear to have been resolved by replacing the RAID shelf on 23rd September 11. A major air-conditioning outage at BlueIce at Swansea in early September has resulted in a suspension of all processing for 4 weeks. FUB has continued processing MERIS BBDRs and has completed 2008, 2009, 2010. 2008 was the first year processed using the new orbit-based BEAM modules. Errors were found in the processing of global lat,lon mosaics which have taken several weeks to resolve. During this process, significant artifacts have been observed in polar (permanent snow/ice regions) which are believed to be due to MODIS priors. These are currently under investigation.

The Test Product Validation Report draft was submitted on 26th September 11. The completion of the TPVR awaits the return of the BlueIce computer (to update the AOD analysis against AERONET) and the final production of global products. Use of AODs from VEGETATION for the calculation of Blue-Sky Albedos was shown to be successful. All the FLUXnet data from 2004 - 2008 (last year of availability) was provided by JRC and will be included in the TVPR analysis. BSRN sites (apart from SURFRAD) proved to be useless for global validation as none were from tower sites and only 2 were in homogeneous land cover area. Only one of these had any corresponding EO data.

The GlobAlbedo web site has been updated in preparation for the roll-out of the GlobAlbedo products - same web address: www.GlobAlbedo.org and passwords.

Executive summary of progress: 31/05/11 - 30/06/11

Progress has been somewhat slower than anticipated over the course of the last month, largely as a result of unexpected software issues – now resolved. The test code for BBDR to Albedo was delivered to UCL as promised on 24/6 by Brockmann Consult. However, there is insufficient time to process all the data by the original QR proposed date. The Python vs Java code testing now shows one-to-one compliance for the final art of the processing chain. Last minute delays include getting hold of an updated DEM (now resolved) from USGS but ATSR continues to be a major headache. Although parameters have been derived for a significant number of orbits from 1995 - 2010 for the ATSR forward-to-nadir co-registration, progress has been slow with the geocoding improvements. Therefore, the consortium has had to request a delay of QR until late July. QR planning now includes the use of video conference (Peter Muller will be in the USA for the proposed QR period). ESRIN have agreed to this proposal on the understanding that the 120 hour test can be started/monitored remotely, i.e. to demonstrated performance before travel.

Executive summary of progress: 30/04/11 - 31/05/11

Work progressed on the co-registration of ATSR forward-to-nadir views, the identification of issues for the cloud and snow masking and the completion of the processing chain for the BBDR-to-BRDF. Further MODIS data of BRDF (4/10 - 3/11) was collected during the MODIS science team meeting so that the MODIS prior could be extended as well as providing a comparison dataset for algorithm testing. A presentation on the GlobAlbedo project was made to the MODIS science team which was extremely well received.

Executive summary of progress: 31/03/11 - 30/04/11

The GlobAlbedo consortium completed the delivery of the CDR data package and supported CDR, held by video conference on the 29th April 11. All but one delivery has been accepted: ATBD change request. The consortium expects a response from ESRIN imminently.

Executive summary of progress: 28/02/11 - 31/03/11

The prototype product for every 8 days in 2005 is nearing completion with almost all of the VGT and MERIS BBDRs processed by SU and FUB respectively and MSSL processing almost all of these input data to BRDFs. Coding is still being completed on the final merged product of the NoSnow and With-Snow DHR and BHR. ATBDs are being updated and just await the final coding and documentation to be completed. No fAPAR LUT was received by the deadline of 1/4/11 so this component will take place with B. Pinty and collaborators and will NOT be included in the final prototype product.

Executive summary of progress: 31/01/11 - 28/02/11

MSSL has (on behalf of the GlobAlbedo consortium) requested a delay of CDR from the 31/3/11 until 30/4/11. This is primarily to address the slower than anticipated processing of the GlobAlbedo data products. A number of artefacts have been identified and are being addressed. It appears that the updated IDEPIX cloud mask behaves much more like the original test using the MODIS prior (in our project's continuing efforts to remove dependencies on MODIS).

Executive summary of progress: 31/12/10 - 31/01/11

Significant progress has been made on production of the prototype product and the prototype product validation prior to the GlobAlbedo UCM-1 which was held at MPI-Hamburg on the 12th January 2011. The meeting was attended by more than 20 people including 10 users. The meeting was an opportunity to share (with the users and ESA) GlobAlbedo developments (and assumptions) over the last reporting year. The users reported very positive feedback on the project and products.

Users have since been invited to provide feedback by way of an on-line questionnaire. This approach has proved quite successful with 7 respondents as of the end January 11. A reminder e-mail will be sent shortly requesting feedback from the outstanding GlobAlbedo users. These will form the basis of ATBD change request deliverable.

UCM-/1CDR deliverables were provided for UCM-1. A number of RIDs have since been raised by ESRIN. It has been agreed (with ESRIN 30/1/11) that the RID responses will be provided 1 week in advance of GlobAlbedo CDR data product is required to complete the RIDs.

The GlobAlbedo consortium will hold CDR at the end of March 2011.

Executive summary of progress: 15/11/10 - 15/12/10

MSSL has successfully responded to the RIDs raised at PM2. The updated ATBD document (issue 2) was accepted on 30/11/10.

The consortium has continued with preparations for UCM-1 (12/1/11, MPI-Hamburg).

The processing system (blades) at MSSL for the BBDR-to-BRDF/albedo processing continues to be developed. Work continues on the deliverable reports (D7-D10) including tile processing of prototype products.

MSSL is awaiting a response from ESRIN regarding MSSL's request to delay CDR until the end of March 11. This request is as a result of the late delivery of (some) data products, the use of non-optimised prototype software and I/O and the stable performance of the processing system at UCL-MSSL.

Executive summary of progress: 15/10/10 - 15/11/10

Progress this month has been focused on normal project activities in addition to the completion of each of the RIDs raised (by ESRIN) at PM2.

ATBD updates/response was submitted to ESRIN on the 15/1/10. The processing time using the prototype code for the generation of 8-day global products for 2005 was estimated to delay the completion of these global products by 2 months.

MSSL will discuss the programmatic issues associated with this delay during a telecom with ESRIN (schedule for week 46).

The consortium has invited the core users to provide co-ordinates for test tiles (one tile each). These tiles will assist with validation of algorithms.

Executive summary of progress: 05/09/10 - 15/10/10

The main body of work for the current reporting period has been the preparation of documents and materials for PM2. The deliverables included: D-5 (Product Validation Plan) and D-6 (Algorithm Theoretical Basis Document - draft).

PM2 was held at ESRIN on the 20th September 10 (see GlobAlbedo_PO_004_V1_0). A number of RIDs have been raised by ESRIN. RID responses will be provided to ESRIN by early November 10.

A project consortium meeting was held at UCL on the 12th October to discuss planning for the upcoming UCM-1, CDR and closure of RIDs raised at PM2.

Actions have been assigned in order to track progress.

The dates for UCM-1 and CDR have now been agreed by the consortium, users and ESRIN. The meetings will be held at MPI Hamburg.

12/1/11: UCM-1(MPI Hamburg)
13/1/11: CDR (MPI Hamburg)

CDR documentation will be provided to ESRIN one week in advance of CDR.

Data is currently late and discussion are being held with suppliers during week 43-44 (personnel currently on leave).

Executive summary of progress: 05/08/10 - 07/09/10

The GlobAlbedo team is working towards PM2, due to be held at ESRIN on the 20/9/10. The project is generally progressing to schedule. The GlobAlbedo management plan has been updated to reflect the recently agreed 2 month slip. This plan has been issued to ESRIN for comment (July 2010).

Data product deliverables are now nearing completion.

The SDR/BBDR processor (developed by FUB) has been upgraded to include all input types (VGT, AATSR and MERIS) as a second consolidated version, able to process top-of-atmosphere radiance data to BBDR data projected onto the MODIS grid for the subsequent albedo retrieval. The BBDR->BRDF/Albedo processor (developed by UCL) has been completed and tested using MODIS, VGT, AATSR and MERIS BBDRs.

A map gridding tool used by the SDR-BBDR processor has been identified as having a “bug” which means that although all the VGT, AATSR and MERIS data are co-registered they are not co-registered to the MODIS grid. A fix is currently under investigation.

Tests have been performed on assessing the impact of different sensors on retrieval of BRDF/albedo which suggest that neither MERIS nor AATSR have much impact due to their low directional sampling capability (MERIS) and their sparse coverage (AATSR). Unfortunately the assessment of the impact of priors awaits a fix for the mapping issue. There are striping artifacts due to missing lines in the VGT data, which need to be further investigated.

BBDR data for 4 test sites have been processed for MERIS by FUB and for AATSR and VGT by SU using code from FUB. Sensor inter-calibration coefficients have been determined for December 2008 / January 2009 for MERIS vs AATSR. Cloud masked VGT data was received too late to be used in time to be included in the draft ATBD.

Executive summary of progress: 08/07/10 - 08/08/10

The project schedule has been updated to include the 2 month slip (as agreed with ESRIN in June 10 - see annex - 2 of MPR). The GlobAlbedo consortium has used this additional time to progress the work to be presented at PM2. Deliverable D-2 (Technical Specification Document) has been accepted by ESRIN, thus closing all outstanding RIDs from PM-1.

The main areas of concern currently included the delivery of data from Swansea University and the Freie Universität Berlin of sample SDRs for all the 4 tiles selected for demonstration in the ATBD. FUB are waiting for the specifications of the gridding/binning to be used by Brockmann before they can complete processing and delivery of these 4 tiles to UCL.

SU have procured a 40TB raid system for the storage of EO data (ATSR-2, AATSR and VGT).

A draft of the ATBD inputs has been received from UCL, FUB, SU and BC.

First versions of the SDR/BBDR prototypes have been developed at FUB and used to process all the MERIS data (all the ENVISAT orbits over that test site for the time period from 12/2004 - 1/2006) from one of the four test sites requested by UCL to test the albedo retrieval processor.

A preliminary BEAM processor has been implemented at SU to generate a preliminary data set of SDRs for VGT based on the climatological aerosol fields provided by the VGT product.

Text input to the validation plan has been submitted by FUB.

UCL has completed processing 10 years of MODIS data into monthly results at 0.5º grid resolution and showed results of inter-comparisons for selected sites on different biomes and how the results reflected differences in BRDF QC and whether snow/ice was present.

Stage 2 of prior processing has started, taking as an input the results of stage 1, consisting of a resampling from 500m to 1km and the application of a temporal smoothing function using all samples available for the different Albedo parameters. Up to August 6th, 63 tiles have been processed.

A Python class for Albedo retrieval is being developed to allow the calculation of Albedo parameters from broadband directional reflectances and their respective kernels.

Executive summary of progress: 08/06/10 - 05/07/10

The date of the next progress meeting (PM2) has been slipped by a period of 2 months; this is in order to help with planning constraints: availability of processed data products. Suggested revised dates have been provided (Table 7-3) and it is planned that these will be formally agreed between the consortium and ESA at PM#2 (20 September 10). The master schedule is in the process of being updated to reflect the 2 month schedule slip; and will be included in an up issue of the project management plan (before 30/7/10).

There is still one minor clarification necessary to close out the PM1 RIDs raised on the technical specification (Ambiguity over TR-28, see discussion later).

Project web site has been updated.

The Project generally continues to progress to schedule (other than for data product). The 2 month slip will relive some of the pressure of playing ‘catch-up’ now that the data products are becoming available. The deliverables due for PM-2 are on schedule to be completed as the major components of the ATBD are well advanced. The calculation of priors is almost complete after processing several Tbs of MODIS BRDF Collection 5500m tiles. However, the project team recognise that there is a need to demonstrate the algorithms using European data. Therefore the emphasis over the summer is on production of BBAs derived from VGT, MERIS and AATSR for 2005 (the test year) to demonstrate that the algorithms are sufficiently mature to produce a test year by the end of Phase 1 and thence for the full 15 years.

CIF (Capital Infrastructures Fund) monies was recently granted from UCL. This will be used to create a large-scale data processing, data storage and robotic tape back-up system at UCL-MSSL for use in processing 15 years of data. It is proposed that this system will be commissioned over the summer, ready for use in processing the year of data in the Autumn. The systems at SU and FUB are already set-up and ready for their level-1b to BBDR processing.

Executive summary of progress: 05/05/10 - 05/06/10

The GlobAlbedo project has progressed at a satisfactory rate during May – June. Early preparations for PM2 have continued, including the preparation of documentation [SoW]. Formal presentations will be prepared over the course of next 2 weeks.

All consortium members now have a WIKI account and JIRA for developers.
RIDs (PM1) have been completed and issued to ESRIN 9/6/10. An initial draft of the traceability matrix has been supplied and will be updated for PM2 when it will be formally reviewed.

GlobAlbedo technical meetings continue to be held on a monthly basis, most recently 7/6/10 (see project web site for minutes).
The project web site has been updated inline with ESRIN requests.

Executive summary of progress: 05/03/10 - 05/05/10

Progress has been satisfactory over the period April - May, despite the teaming issues caused by the recent volcanic activity in Iceland (all but 1 member of the project have now returned). There is likely to be an impact due to the 4 week delay in return by P. North from Australia, particularly with regard to SPOT-VGT ATBD. There was also an impact on completion of the RIDs and the ACTION items outstanding but these are both likely to be resolved very shortly after everyone’s return.

Data continues to be collected (as planned) and data transfer speed tests have been completed between MSSL and members of the consortium – few minor areas to be reviewed.

In addition to normal programme developments, a BEAM programming course was held at BC during late April. The course was aimed at training the GlobAlbedo development teams (UCL/SU/FUB) to develop the GA processors. A software development team meeting was held immediately following the BEAM programming course, in order to better define the detailed responsibilities for: ATBD, breadboarding and software.

It is expected that invoices (KO advanced payments) will be processed in the coming week – delayed by updates to ESA SAP (replacement of EFIS). Internal technical progress meetings continue to be held on a monthly basis, TM#3 was held on 11/3/10 and TM#4 on 19/4/10 on SKYPE telecon. There are no major issues or concerns at this time.

Executive summary of progress: 05/02/10 - 05/03/10

GlobAlbedo progress is generally progressing as expected; and without major issue. This is despite the delays in processing the main contract.

RIDs were received for BR and TS (PM#1); which have subsequently been distributed to the consortium team. Responses will be provided to ESRIN in March 2010.

All SPOT-VGT2 data for the period 2003 - 2008 has now been delivered on LTO-3 tape and read onto RAID array disks at SU. The outstanding data (VGT-1 and 2009 VGT-2) has been requested. 1.5Tb disks have been used to copy the majority of the 10 years of MODIS Collection 5 albedo product data.

Discussions were held at NASA Goddard Space Flight Centre after the MODIS team meeting about possible collaboration on developing a common protocol with CEOS-LPV and the MODIS land validation activities. It was agreed to pursue this both with the GlobALbedo consultants (Schaaf & Schaepmann-Strub who are co-chairs of the Albedo focus group within CEOS-LPV) and with colleagues at GSFC, Ascii data has been prepared of all the BSRN, SURFRAD and tower albedo measurements which will be provided to GlobAlbedo during March. Processing system development continues.

Executive summary of progress: 05/01/10 - 05/02/10

Progress meeting #1 was held at Météo-France on the 18th January 2010. The meeting was supported by users from Météo-France and MPI Hamburg. The consortium provided detailed presentations which can be found on the project web portal, as can the minutes of meeting. The PM#1 presentation included a review of the requirements baseline document (SoW D-1) and technical specification (SoW D-2) document. ESA have raised a number of minor RIDs (5) on the requirements baseline document which will be addressed by the consortium over the course of the next 2 weeks. A number of additional RIDs (13) were raised on the technical specification document which are currently being processed.

Just over half of the MODIS 10 year archive of Collection 5 albedo products were returned by JPM on 30/1/10 on 1.5TB USB_2 disks from NASA GSFC. It is planned that the remainder of these data will be brought back on 5/3/10. In addition, the Long-Term Data Record (LTDR) of SDRs at GAC (4.4km resolution) will be brought back at this time. JPM gave a well received general presentation on the GlobAlbedo project to the MODIS Land team in Washington DC on 26/1/10. He also discussed arrangements with the prospective GlobAlbedo consultant, Crystal Schaaf, about transfer of screened global BSRN data that had been employed previously for validation of MODIS albedo products which will take place in the next few months. The revised ESA contract has now been received at UCL-C and sub-contracts are currently being issued.

Executive summary of progress: 05/11/09 - 05/12/09

A successful Kick-off meeting was held at ESRIN on 5/11/09 with the GlobAlbedo team and 6 users in attendance [AD1]. A subsequent Technical Meeting (TM#1) [AD3] was held at MSSL on 25/11/09, during which a number of questions were formulated based on the user requirements described at the KO and the TRs in the ITT. These clarifications of URs are being incorporated into a questionnaire for completion by the 6 users for the Baseline requirements document; and any other users identified. TM#1 also detailed some of the Technical Specifications needed for PM1. JPM has continued to try to identify users in the UK Met office and they have agreed to provide user input to the questionnaire. However, they are unable to provide a user to attend future meetings. ECMWF continually fail to respond to requests. JPM is in correspondence with KNMI to find a suitable user.

One of the users, Jean-Louis Roujean of Météo France kindly offered facilities to hold the PDR/PM1 at his premises on Monday, 18/01/10. JPM has suggested that if team members were able, that a meeting should be held in the previous afternoon. UCL has proposed amendments to the ESA contract which are being dealt with and subject to contract agreement and signature will be able to issue sub-contract and consultancy agreements within 2 weeks. JPM has been in discussion with NASA GSFC MODIS and NASA Langley DAAC about obtaining a copy of a decade of the latest processed MODIS and MISR products. It is hoped that JPM will be able to return with a LT04 tape copy of the entire MOD43 albedo product when he attends on GlobAlbedo's behalf the next MODIS science team meeting (21st - 29th January 2010). JPM met with the MISR LaRC Manager at a recent MISR team meeting and it was agreed that LaRC would subset the global MISR 2AS product to extract parameters such as BRFs and narrowband and broadband albedos. This dataset should be <6TB so would be FTPed. An analysis would also be conducted of any trends and an assessment is proposed of whether RTLS could be applied to the entire spectral BRF dataset as a test for future GlobAlbedo products. MSSL has provided LT04 tapes of SPOT-VEGETATION L2 data to Swansea University (25/11/09). This is in order to enable SU to investigate data format issues whilst BC write a reader for VGT products as part of BEAM. JPM submitted 2 GlobAlbedo abstracts (one on the project and the other lead by DP on his study for QA4EO) to the ESA “The Living Earth” symposium and plans to submit an abstract to IGARSS 2010 by 18/12/09 as requested by ESA.

Upcoming Events: QR will be held September 2011

Current Activities: Preparations for QR

Home

Site Map

Disclaimer

Accessibility

Privacy

Advanced Search

This website passes W3C HTML Quality Assurance