WMO SPACE PROGRAMME Workshop on Operational Space-based ...€¦ · WMO SPACE PROGRAMME Workshop on Operational Space-based Weather and Climate Extremes Monitoring Geneva, Switzerland

W O R L D M E T E O R O L O G I C A L O R G A N I Z A T I O N

==============================================================

WMO SPACE PROGRAMME

Workshop on Operational Space-based Weather and Climate Extremes Monitoring

Geneva, Switzerland

15-17 February 2017

MEETING REPORT

WMO General Regulations Regulation 42 Recommendations of working groups shall have no status within the Organization until they have been approved by the responsible constituent body. In the case of joint working groups the recommendations must be concurred with by the presidents of the constituent bodies concerned before being submitted to the designated constituent body. Regulation 43 In the case of a recommendation made by a working group between sessions of the responsible constituent body, either in a session of a working group or by correspondence, the president of the body may, as an exceptional measure, approve the recommendation of behalf of the constituent body when the matter is, in his opinion, urgent and does not appear to imply new obligations for Members. He may then submit this recommendation for adoption by the Executive Council or to the President of the Organization for action in accordance with Regulation 9(5).

EXECUTIVE SUMMARY The Workshop provided for a dialogue amongst satellite operators, WMO Regional Climate Centres (RCCs) hosted by National Meteorological and Hydrological Services (NMHSs), and the science community to stimulate the utilization of space-based observation data and products for monitoring selected weather and climate extremes (heavy rainfall and drought in particular) on a routine basis (“in operations”), in response to current and future user requirements.

_________________________

From bottom row left:

Yuriy Kuleshov, Riris Adriyanto, Elena Manaenkova, Carolin Richter, Mark Dowell, Tillmann Mohr, Toshiyuki Kurino, Yang Jun, Stephan Bojinski, Peter Salamon, Ali Behrangi, Joerg Schulz, Ladislaus Changa, Pingping Xie, Kiyotoshi Takahashi, Andre Obregon, Einar Bjorgo, Riko OKI, Rainer Hollmann, Ralph Ferraro, Simon Eggleston, Donald Hinsman

1

1 SETTING THE SCENE 1.1 Opening remarks Mark Dowell opened the workshop at 9.00 on Wednesday 15 February 2017.

Elena Manaenkova, Deputy Secretary-General of WMO, welcomed participants. She associated herself and WMO management with the Space Programme and stressed its importance. She noted that Members and UN agencies are putting high priority on strengthening climate resilience and early warning systems by all Members. Monitoring and forecasting are key assets of the WMO community that are needed for building such systems. Tools are available through observations, provided on the ground and from satellites. GCOS has defined the ECVs, the WMO State of the Climate reports have become an important product, and there is more demand for producing impact-based statements and forecasts. The modelling community through WCRP is ready to inform on likelihood of extreme events under a changing climate, but often the information is not availability sufficiently quick and readily useable. 1.2 Introduction of participants The provisional agenda was adopted. Participants introduced themselves in a tour-de-table (see Appendix A). 1.3 Objectives of the workshop T. Kurino introduced the definition of “extreme weather event” and “extreme climate event” according to WMO Meteoterm, adopted from the IPCC AR5 WG I report, as follows: “An extreme weather event is an event that is rare at a particular place and time of year. Definitions of rare vary, but an extreme weather event would normally be as rare as or rarer than the 10th or 90th percentile of a probability density function estimated from observations. By definition, the characteristics of what is called extreme weather may vary from place to place in an absolute sense. When a pattern of extreme weather persists for some time, such as a season, it may be classed as an extreme climate event, especially if it yields an average or total that is itself extreme (e.g., drought or heavy rainfall over a season).” (WMO Meteoterm; IPCC AR5 WG I Glossary). The four objectives of the workshop were:

(1) To present use-cases for satellite-derived products focusing on monitoring extreme events with regard to ‘accumulated high precipitation’ and ‘drought’ (“current provisions baseline and practices”)

(2) To understand and document the requirements from WMO Regional Climate Centres (RCCs) / other climate centres for monitoring weather and climate extremes in operations, with a focus on ‘accumulated high precipitation’ and ‘drought’ ; these requirements are not necessarily related to satellite data – many RCCs currently do not use space-based data (“current needs, requirements, practices”)

(3) To identify lessons learned in the formulation of weather and climate extremes-related requirements and satellite-specific responses, and to formulate a preliminary response by RCCs/climate centres to the identified requirements (“looking forward - matching user needs and provisions”)

2

(4) To assess the feasibility of a demonstration of space-based weather and climate extremes monitoring in operations, for strengthening the capacity of NMHSs (“looking forward: demonstration activity”)

For the purpose of this workshop, participants specifically looked at:

users requirements for monitoring weather and climate extremes on a timescale of pentad (5-day) to weekly up to monthly basis

satellite-based products available on near-real-time basis for monitoring “heavy precipitation” and “drought” events on a routine basis (“operationally”) for climate analysis and monitoring, and for the development of improved climate services.

He showed the use of a satellite-derived precipitation product over South-East Asia (Lao PDR) and its limits in detecting heavy rainfall. The challenges to be addressed include:

To calibrate long (>10 years) time-series of satellite data using in-situ data, and derive percentiles;

To validate short time series of satellite data and products. Participants noted that stakeholders should not be limited to RCCs but also other institutions involved in climate services. 1.4 Architecture for Climate Monitoring from Space Mark Dowell introduced the Architecture for Climate Monitoring from Space, with details on the basic concepts, and on the 2015 WMO-GFCS-EC case study report1. The baseline requirements for this process come from the GCOS-led cycle of adequacy and implementation reports. To highlight the impact of this baseline on programmatic decisions, he noted that an estimated 200-250m € of investment made in European institutions dedicated just to ECV production has been possible thanks to the existence of GCOS plans. The Architecture provides a framework in which (i) it becomes visible to users of ECV records what assets exist, (ii) gaps and missed opportunities are identified, to enable remedial action, (iii) coordinated satellite mission planning is facilitated. The functional flow (“logical Architecture”) is traceable to GCOS monitoring principles and guidelines. The ECV inventory, to be made available in mid-2017, will be a first instance of the “physical” Architecture, in line with the 4-year roadmap of the CEOS-CGMS Working Group Climate. In the Architecture, the precise definition of ICDRs is missing, and the workshop should provide further insight. The WEF Global Risks Landscape 2017 identified extreme weather events as the top-rated risk in terms of combined likelihood and impact.

1 Satellites for Climate Services : Case Studies for Establishing an Architecture for Climate Monitoring from

Space, WMO-No. 1162, https://library.wmo.int/pmb_ged/wmo_1162_en.pdf

3

2 EXPECTATIONS TO OPERATIONAL SPACE-BASED WEATHER AND CLIMATE EXTREMES MONITORING

2.1 GEO Barbara Ryan briefed on the mission of GEO, updated on the number of Members (104) including the European Commission and 106 Participating Organizations, including WMO, GCOS, GTOS and GOOS. The new global work plan to implement the Global Earth Observation System of Systems (GEOSS) 2016-2025 is now in force. Director Ryan recalled the cross-cutting role of climate and weather in GEO’s Societal Benefit Areas (SBAs) and highlighted the engagement priorities 2017-2019 (Sustainable Development Goals, Climate Change, Disaster Risk Reduction). GEO works to provide and promote open EO and geospatial data to monitor and achieve the SDGs. A mapping of ECV products against these indicators is recommended. GEO has an advocacy role in getting UN member countries and custodian agencies for individual targets to harness the potential of EO and geospatial data for improved policy making. In supporting Disaster Resilience, space agencies are collaborating using the joint CEOS-CGMS Working Group Climate as a model for collaboration. GEO-DARMA and GEO-GLOWS are activities in this regard. The Global Drought Information System (GDIS) and the Global Flood Awareness System (GloFAS) , both activities in the GEO Work Programme, were presented. WMO is a contributing institution. The GEOSS Common Infrastructure (GCI) offers a brokering service to a large amount of data providers, which includes the WMO Information System. Director Ryan highlighted the Australian Data Cube application as an example for mining a long time series of satellite imagery for information on flooding and drought. Similar applications could be developed for monitoring change in forests and urban areas. 2.2 GCOS Carolin Richter briefed the workshop participants on the latest GCOS status reports and plans. The new 2016 implementation plan describes the global needs for having an enhanced global observing system for climate. Climatecouncil.org.au has a finding that generally, extreme heat waves and cold spells are high-impact consequences of climate change. She made a plea that recommendations from the workshop should also consider needs related to surface-based observing networks (density, data availability, data formats). She quoted actions in the 2016 plan related to temperature, precipitation, and other variables (sea state, ocean surface heat flux, ocean surface stress, river discharge, snow parameters). For adaptation, “even the smallest pixel is too large”, and higher spatial and temporal resolution is required by many users. Current development of historical climate indicators envisages six areas: temperature and energy, atmospheric composition, oceans, cryosphere, land use/vegetation change, extremes heat waves. They are intended as a communication tool for WMO and GCOS, and satellite operators will have an important role in the provision of supporting data.

4

She made reference to Annex A in the 2016 GCOS plan, which specifies ECV product requirements that are a priory technology-agnostic. Refinement of these requirements is planned for the coming two years, in consultation with the community. The new GCOS requirements for ECVs have been reflected in the space agency-led ECV Inventory, as far as this was possible given the timing of both efforts. It was recognized that many NMHS impose restrictions on the redistribution of gauge-based precipitation observation data that they submit to GPCC, limiting the provisions by GPCC to products (not raw datasets). 2.3 GFCS Filipe Lucio stressed the user-driven nature of GFCS in all its pillars, with focussed action in five priority areas. More than 70 countries have only basic climate services or even less, and these are subject to targeted action. He summarized gaps and deficiencies identified in the GFCS implementation plan. He suggested whether satellite data can be used to derive proxies that help infer on socio-economic information (LAI, NDVI as proxy for drought). Among the eight key climate information needs identified by users, many are concerned with assessing vulnerability and decision-making at local levels. Impact-related datasets such as on crop yields, river flows, health statistics are in greatest need. Satellite data may contribute to building observational databases to address biophysical and socio-economic variables. In October 2016, the IBCS management board approved priority needs for operationalization of the GFCS in the 2017-2019 period, along three main lines. Linkage of RCCs and agricultural services are very strong in Tanzania. Agriculture is a top priority for sustaining livelihoods and within national climate change adaptation plans. Someone from the NMHS is embedded in the agriculture extension service, and agriculture experts are part of NMHS projects. 2.4 WCRP Boram Lee briefed on WCRP and its role in understating the range of scales of climate variability and change (week – season – decade – century), phenomena and tools (observations, models) that are of interest to the various WCRP projects (CliC, CLIVAR, GEWEX, SPARC, CORDEX). Seven Grand Challenges have been identified to further understanding and predicting the climate system and its changes. Currently the WCRP Grand Challenge on Understanding and Predicting Weather and Climate Extremes (GC-Extremes: https://www.wcrp-climate.org/grand-challenges/gc-extreme-events) addresses critical gaps exist in amount, quality, consistency, and availability of data, especially for extremes. These are associated with undigitized datasets as well as lack of data exchange. Recent studies also highlighted shortcoming in our ability to characterize not only modelled but also observed extremes. For example, over many land areas, precipitation intensities demonstrate a large disparity in CMIP5 and a systematic over-estimation of wet days, and no clear agreement with and between in-situ and remotely sensed datasets. The Grand Challenge-Extremes also emphasized the recent improvement in documenting and understanding climate extremes by coordinating in situ, remote sensing and reanalysis data (e.g. CLIMDEX), and called for enhanced support for the relevant activities of WCRP and CCl through the joint Expert Team on Climate Change Detection and Indices (ETCCDI).

5

Through cross-domain and cross-programme activities, research under the Grand Challenge-Extremes has made progress in the analyses of interaction between large-scale and regional-scale phenomena, as well as land-atmosphere feedback and forcing mechanisms; e.g., global scale drivers for regional/local extremes such as the Pakistan flood in 2010, and strong regional variability within the global water availability projection (“dry gets drier, wet gets wetter”) associated with land-atmosphere feedbacks. Another example from 2010 is the Indus valley flooding and Russian heat wave connection due to a blocking pattern. Participants also noted that there is varying skill of models in simulating extreme events, and that issues are different at various scales. High-resolution models are critical to improve forecasts of short-lived extremes (heavy rainfall event); land processes are a constraint to improve modelling of more long-lived extremes (heat wave). As to the potential to predict extremes from/through remote sensing: monitoring of weather patterns established 1-2 months in advance of an extreme event, such as pressure patterns, may provide additional skills combined with good understanding of scale interactions. One also needs to look at “compound extremes”, e.g., the connection of heat wave and flooding event in different locations. In this context, the workshop took note and welcomed the Grand Challenge-Extremes’ focus on compound extremes, and looked forward to the publication of a high-impact overview paper on climate extremes. The WCRP Open Science Conference in 2018 on Extremes and Water Availability focuses on two Grand Challenges. The connection between the Grand Challenge-Extremes, and SDGs was highlighted. The GC-Extremes recommended that further investigation could be made toward an improved agreement/synchronization and integration of in-situ observations and satellite products (e.g. precipitation), possibly on common space and timescales. It also recommended to support extended use of new systems (e.g. GPM), and to improve datasets on various parameters such as soil moisture. 2.5 CLW/DMA Peer Hechler represented the WMO Commission for Climatology (CCl) and the World Climate Services Programme (data and monitoring part) in support of operational climate monitoring activities by NHMSs. He briefed on the WMO Statements on the State of the Global Climate2. While relying on national contributions and strongly informed by ground-based observations, satellite data contribute to statements on ocean SST, heat content, sea level, sea ice cover. As a matter of fact, WMO climate monitoring relies basically on national in-situ data and assessments for historical reasons, especially as satellite data became available only from the 1970s onwards with the ability to generate robust time series since the 1990s. Accordingly, satellite-based data have been introduced in WMO climate monitoring activities over the last twenty years. Key chapters of the WMO Annual Statement include (reflecting key data needs):

Global temperature analyses

Global precipitation and snow cover analyses

Oceans (SST, heat content, sea level)

El Nino including global impacts

Cryosphere (sea ice extent, Greenland melt area, etc.)

Regional extremes (particularly heat- and cold waves, heavy precipitation/flooding, drought/wild fires, extremes and records)

Tropical cyclones (number, track, impact)

2 http://www.wmo.int/pages/prog/wcp/wcdmp/CA_2.php

6

Greenhouse gases (status, trends)

Stratospheric ozone/ ozone-depleting substances (status, trends) Required additional content of the Statement is mostly around impact information (impact of climate anomalies and extremes). WMO Regional Climate Centres (RCCs) have mandatory functions for several domains, including for climate monitoring and climate data: (i) climate diagnostics (climate variability and extremes at regional and national scale; mean and maximum values of), (ii) establishing a reference climatology, (iii) establishment of climate, quality-controlled regional climate datasets, and (iii) implementation of a regional climate watch. Climate Watch3 describes a concept that provides advisories on ongoing and expected anomalies with potential adverse impact on society, i.e., heat wave, drought, hot and cold spells. A Climate Watch system consists of pillars: climate data, climate monitoring, long-range forecasting, communication. Climate Watch advisories are a national responsibility with a high potential for RCCs to provide regional guidance information. Climate watch formats vary depending on climatic zones. WMO runs regional climate watch implementation workshops. Climate Watch data and monitoring requirements depends on regional/national vulnerabilities and risks, and hence are different for different regions (e.g. cold spells/icing in the far North and South; sand storms, monsoon characteristics/precipitation, cyclones in the Tropics; sand storms in West Asia; droughts for drought-prone areas etc.); regional Climate Watch implementation workshops help identify priorities. Satellite data and products are used already for WMO climate monitoring activities, however, to extend their use to cover areas with sparse or missing in-situ observations and/or to introduce new variables and indices, the following requirements have to be considered: - Robust data sets, well maintained, quality-controlled (homogenised where feasible) and documented, that span ideally 30 years (for context information/climatology), and that are likely to be maintained in the future (a CCl task team is investigating the ECV inventory accordingly) - Capacity development for NMHSs to enable the most efficient use of satellite data for national climate monitoring, analyses and assessments - High-resolution data sets based on robust methodology to enable time series analyses including extremes identification - Availability of quasi real-time products to feed daily/weekly/monthly operational climate monitoring activities - Integrated data sets; improved coordination of in-situ, remote sensing and reanalysis communities The short discussion revealed the need to develop integrated guidelines for Members to efficiently apply in-situ and remote sensing (and reanalysis) data including addressing the different data characteristics (station data vs areal data). References: Use of satellite-based products in WMO operational climate monitoring activities, particularly - WMO Annual Statements on the Status of the Global Climate (and aggregated five and ten years statements); additional reading cf. http://www.wmo.int/pages/prog/wcp/wcdmp/CA_2.php

3 http://www.wmo.int/pages/prog/wcp/wcdmp/CWS_1.php

7

- WMO Regional Climate Centres including Climate Watch systems; additional reading cf. http://www.wmo.int/pages/prog/wcp/wcasp/rcc/rcc.php and http://www.wmo.int/pages/prog/wcp/wcdmp/index_en.php (-> Climate Watch system) Participants enquired how the feedback loop between the Architecture effort and WMO guidance material on climate can be improved (regarding new techniques, new application areas)? This is a task of the CCl Task Team on the use of remote sensing and satellites. A shift in community practices how to detect extremes may be warranted (point-based vs areal average-based). 2.6 CLW/Agricultural Meteorology José Camacho briefed on needs of the agricultural meteorology (“agrimet”) community. Key questions are (i) risk identification (weather and climate factors), and (ii) impact assessment on agricultural production. Decision-making at strategic and tactical level require different types of information. Monitoring drought conditions is of primary importance to agriculture. Agrimet services usually rely on historical data, experience with crop types etc. A crop model is developed from which advice to farmers is derived. He quoted an example for rainfall-based crop planting advice in central Mali, using plot-based rain gauge and remote sensing. WMO inter-Commission activities aim to improve quality and traceability of rain gauge measurements. For deriving soil moisture, ground-based and satellite data (MODIS) have been used. From remote sensing, key parameters for agricultural applications are rainfall estimates, evapotranspiration, cloud cover, radiation, vegetation condition indexes, soil moisture, wildfire risk indexes, and land use, cloud masks, and snow cover area. Many EUMETSAT Land-SAF products are being used. He showed a decision-making cascade based on seasonal forecasts, 10-day forecasts, and short-range forecasts with respect to the agricultural community. He asked whether or not Land-SAF–type products were produced systematically, also by other communities (ACMAD, EC-JRC, FEWS-Net)? A regional approach will lead to definition of specific needs of the agricultural meteorology community (data, products). 2.7 European Commission Copernicus Emergency Management Service Peter Salamon briefed on the Copernicus Emergency Service. He stressed a number of important issues on the use of operational space based weather products that should be addressed. This included (1) the increase of temporal and spatial resolution of space-based products where technically possible as forecasting and monitoring of weather extremes usually require high spatial resolutions and frequent revisit times; (2) availability of harmonized, inter-calibrated long time series to derive anomalies and enable inter product comparisons; (3) the necessity for uncertainty estimates in the spatial domain of space-based data, such that users in various geographical areas get an idea on the magnitude of errors (e.g., in percentage, level of confidence, probability) and to facilitate the assimilation of those products. Carolin Richter suggested to build upon a future global surface reference network. A middle layer between providers and end users is needed to help users in the selection of products, since there will always be many. Who could this be?

8

The International Precipitation Working Group (IPWG) goes some way in providing this service to users. 2.8 UNITAR/UNOSAT

Einar Bjorgo presented an overview of UNOSAT fully dedicated to satellite imagery analysis and capacity building, servicing UN agencies and UN Members. UNOSAT has four offices (Geneva, Ndjamena, Nairobi, and Bangkok). UNOSAT provided disaster rapid mapping activations, working together with the Int’l Charter Space and Major Disasters, using free and open data, and sometimes purchased data. Examples showed flood extent mapping in Myanmar using radar imagery,. On their flood portal, UNOSAT use archived cases for disaster risk reduction. Often, flood mapping is limited by the timely programming of satellite acquisition. The Flood Finder model maintained by CIMA (Italy) provides a forecast capability to floods, with a view to anticipate the need for imagery. Information on vulnerability and exposure is also included in UNOSAT products. Capacity building requires identification of needs, data access for trainees, repeat trainings, and technical backstopping. UNOSAT works with institutions to enable them access, analyse and present geospatial information, such as climate outlooks produced by IGAD-ICPAC. Another effort looks at mapping water resources in Chad. He argued that many information sources exist, but are often not directly available to users. Sometimes products do not have to be perfect to be useful. UNOSAT provides an extension and outreach service. Ali Beranghi highlighted the challenge of identifying the needs of end users of satellite-based products – a broker such as WMO is needed between providers of satellite data and end users. More effort is needed to work on the interfaces between end and intermediate users, and other interfaces in the value chain. Intermediate data users are essential in the value chain, as a “transmission belt”. More focus on identifying local needs is necessary, since there are relatively few users of “global products”. Co-development and training are key components, since many end users do not exactly know what they need. This finding was confirmed in the Case Studies report in support of the Architecture for Climate Monitoring from Space. WMO World Weather Watch is based on global, regional and national forecasting capabilities, with defined interfaces. The GFCS is conceived by the same principle, and it will take time to develop interfaces and document user needs at the various levels within its User Interface. Erica Allis suggested that this may be a topic for the GFCS Partner Advisory Committee; leveraging local universities will be an important component of building capacities in countries.

9

3 CONTRIBUTIONS OF INT’L SCIENCE GROUPS TO SPACE-BASED WEATHER AND CLIMATE EXTREMES MONITORING

3.1 International Precipitation Working Group (IPWG) Ralph Ferraro briefed on IPWG status (participation approaching 500 experts) and activities. From its inception in 2002, some accomplishments include the presence of 166 and 183 GHz channels on GPM, IODC continuity of service, and utilization of post-operational satellites. The recent 8th workshop of IPWG was held jointly with the international snow science group. The Group makes recommendations to CGMS and responds to actions placed by CGMS. IPWG maintains dataset listings: as examples, he highlighted the TRMM 3B42 climatology, and land areas affected by TCs, and extremes in rainfall (95th percentile). He highlighted a global flooding product (Adler et al.), and work on understanding regional biases in algorithms depending on types of precipitation. A large sample of data is needed to examine this, specific Z/r ratios are needed depending on precipitation types. There are many instances were satellite precipitation outperforms NWP models; he cited some well documented cases from Southern Italy, as examples. IPWG members are looking at new ways to improve satellite precipitation, for example, methods to combine different sources of satellite data (Turk et al.): colder temperature in 89GHz against 10GHz background works well as indicator for rainfall. Additionally, progress is anticipated from space based lightning sensors; using cloud surface temperature and lightning (Wang et al.) combined to infer on rainfall rate. IPWG members maintain several regional validation sites; exploitation of these statistics as well as expansion to other domains will lead to algorithm improvements and error characterization. Knowledge of errors and uncertainties associated with products are very important for operational users (P. Salamon, EC JRC). However, the algorithm developers need specific error metrics from the users as these may vary for different applications; extreme events may focus on errors on the high range of rain rate whereas as longer term climate may find bias as the most important parameter. A question was raised whether there is a need for any targeted development of region-specific algorithms or products; whether systematic analysis of extreme events and performance of the various products is required. 3.2 WCRP Global Energy and Water Cycle Experiment (GEWEX) Jörg Schulz presented on GEWEX mission and activities related to climate extremes. The mission is to measure and predict global and regional energy and water variations, trends and extremes. GEWEX consists of four scientific groups (GASS – cloud process studies, GLASS, GHP – regional projects, in-situ observations, GDAP – guiding production and evaluation of products). Data assessments characterize strengths and weaknesses depending on the application, including extremes. GPCP is an example for an activity producing ICDRs 10 days after the end of each month. Integrated data activities aim at collecting datasets on common space and timescales. He quoted IPCC AR5 statements on extremes. Space and timescales of weather and climate extremes vary – and observing systems addressing these will look different depending on the scale.

10

He quoted an analysis of most “intense” storms based on lightning rates, and rainfall rates (Zipser et al., BAMS paper). A TRMM-based result is to estimate the contributions of mesoscale convective systems to tropical precipitation (5% of systems responsible for 50% of total rainfall). Organization of convection and frequency of heavy rainfall is strongly dependent on vertical wind shear, and on surface temperature. A study by Tan et al. (2015) looking at the weather state investigated “organized tropical convection”, which plays a dominant role in tropical rainfall, can be seen by satellites, but is largely missed by models. He showed typical assemblage scales in space and time using satellites over Europe with skill scores. Weather radar data should be exploited as well – there is no concerted effort as yet, due to issues with data availability and sharing of algorithms. He alluded to the finding of stronger warming over grassland than forests, and the relationship with likelihood of rainfall over wet and dry areas. Drought onset and end are hard to predict, and monitoring thereof is not straightforward either. Reprocessing – it is yet to be determined how often it should be done, with what latency, to satisfy user needs. 3.3 EUMETSAT Climate Monitoring Satellite Applications Facility (CM SAF) Rainer Hollmann provided an overview of CM SAF (www.cmsaf.eu) activities and some of the basic concepts (FCDRs, TCDRs, ICDRs). CM SAF provides climate data records on different variables and on a range of spatial scales. There are a number of accompanying activities to support CDR production and reprocessing of satellite data records. A MWI-based FCDR is being produced, and a global precipitation product is planned within the next five years. He presented first results where the DAPACLIP precipitation climatology has been used to generate a global ETCCDI climatology. These results have been published in Dietzsch et al. (2017). CM SAF aims to investigate whether trends in CCDI based on satellite data can reproduce results by Alexander et al. (2006). He presented the CM SAF definition of an ICDR (Interim Climate Data Record) and stated that an international agreement of definition of ICDRs is needed. He noted that real-time products always become better suitable for climate monitoring applications, not only when using GSICS corrections. The main question is then what the delta is between NRT products and ICDRs? Who is responsible for producing ICDRs? It is clear that ICDRs may not be fully consistent with background CDRs as the underlying calibration (i.e., an FCDR) will not be available with the needed short time latency. 3.4 Sustained Coordinated Processing of Environmental Data Records for Climate

Monitoring (SCOPE-CM) Jeff Privette provided an overview of SCOPE-CM mandate and growth opportunities. Four strands of activities are:

- Continue emphasis on maturing CDR products - Emphasize transition to sustained production of CDRs - Support climate monitoring communities - Focus on coordination and sustainment

11

SCOPE-CM provides an opportunity to synchronize production of precipitation CDRs among various centres. A question on how to improve physical consistency of CDRs was raised. Usually this requires a specific effort by investigators to analyse; e.g., within the ESA CCI, where the climate modelling user group looked at ocean SST, ocean colour and sea ice records for investigating changes in Arctic waters. 3.5 CCl/WCRP/JCOMM Expert Team on Climate Change Detection and Indices Ali Behrangi (NASA JPL) introduced the CCl/WCRP/JCOMM Expert Team on Climate Change Detection and Indices (ET-CCDI). It has a focus on the development of a set of indices (currently the set includes a total number of 27 indices) and software that computes indices that focus primarily on climate extremes. Guidelines on the analyses of extremes using these indices were published in 20094. Indices are used to provide a scientific contribution to observing trends, to evaluate models and make climate projections, and for detection and attribution studies. Ocean-related indices and runoff-related indices are currently being developed. Satellite data should be included in the fold (not yet the case). A 2015 workshop 5 of the Team agreed on a 3-year work plan to investigate the role of observations to detect and analyse weather and climate extremes, including satellite-based quantitative precipitation estimates and 3D structures of rainfall events. Herold et al. (2017)6 in JGR-Atmospheres looked at use of remote sensing data to derive indices. Products sometimes show significant differences at different scales for different indices, and the spread among products increases with higher resolution. Some indices are more sensitive than others to spatial resolution. The data are available to make region-specific analyses and to trade resolution against accuracy. The range of errors in observational uncertainty can be significant as highlighted in a study by Herold et al. (2016) in GRL which showed that for a measure of daily precipitation intensity (the SDII index recommended by ETCCDI), the spread across observational products over land was comparable to the spread across climate models. Sub-daily processes are being investigated in a European INTElligent use of climate models for adaptatioN to non-Stationary climate Extremes (INTENSE) project (Fowler et al.). This effort will likely generate extreme indices, some of them based on ETCCDI. INTENSE is collecting a global dataset of sub-daily precipitation data from national met agencies with which to generate indices and analyse the processes causing sub daily extremes. Sub-daily data are important for process understanding and for capturing precipitation peaks and streamflow simulations, spatial at basin scale. One suggestion is to use a collaborative model such as SCOPE-CM to generate climate indicators

4 Albert M.G. Klein Tank, Francis W. Zwiers and Xuebin Zhang, 2009: Guidelines on Analysis of extremes in

a changing climate in support of informed decisions for adaptation WMO-TD No. 1500 5 https://www.wcrp-climate.org/index.php/extremes-data-wkshp-about

6 Herold, N., A. Behrangi, and L. V. Alexander, 2017: Large uncertainties in observed daily precipitation extremes over

land. Journal of Geophysical Research: Atmospheres, 122, 668-681.

12

4 USERS PERSPECTIVES OF WMO RCC AND NMHS 4.1 NOAA/CPC Pingping Xie introduced applications of satellite data at NOAA/CPC, addressing a range of CDRs (SST, OLR, precipitation). These are used for monitoring and diagnosing climate variability and change. Satellite data generated by satellite centres are often not readily usable by climate centres, due to lack of reprocessing, or delays in production. CPC integrates several raw satellite datasets into their climate analyses, some directly from NOAA/NESDIS. OLR datasets are used for monitoring of ENSO, MJO and other phenomena of inter-annual and sub-seasonal variability, and for deriving precipitation estimates. Blending information from multiple sources is an effective way to exploit the strengths of each individual data type. He noted that model forecasts tend to outperform satellite-based precipitation estimates in mid-latitudes during cold seasons, and vice versa (i.e., satellite estimates outperforming model forecasts) in tropical regions and for convective precipitation. He introduced the CMAP and CMORPH datasets, and bias correction of the CMORPH satellite precipitation estimates using gauge-based analyses and PDF-based approaches (satellite-derived estimates tend to over-estimate rainfall intensity at low absolute values, and under-estimate at high absolute values). The second generation CMORPH being developed will take into account snowfall. Based on CPC’s experience with satellite data applications, Dr. Xie emphasized that climate users need long-term records (usually more than 30 years) with reasonable temporal homogeneity, and capable of capturing precipitation of various intensity to ensure fidelity of the probability density function (PDF) structure. Fusion of information from multiple in situ and space-based sources is often required to achieve the goal. CPC currently does not consider inhomogeneities in the level 1 data when blending several level 2 data sources. This may have implications on the quality of the blended product. 4.2 JMA/Tokyo Climate Center Kiyotoshi Takahashi presented activities of the Tokyo Climate Center (TCC). It serves as WMO RCC in RA II since 2009, and provides climate data and products, and delivers capacity development activities. The products include seasonal products, El Nino outlooks, reports on extreme events, global warming, and other data types. These are delivered to NMHS for use in national contexts. He quoted the example of the August 2016 Lao PDR heavy rainfall event, for which the heavy precipitation index exceeded 100% (based on SYNOP and CLIMAT station reports), as shown in the weekly report on extreme climate events. Total accumulated precipitation was estimated at >1000mm, but gauge stations recorded a maximum of only 380mm. He pointed out the JRA-55 reanalysis, including a real-time analysis after 2013 which has a latency of two days. There is also a JRA-55C reanalysis only using surface-based data, which could be useful for comparison with satellite-derived products. Different criteria are applied depending on the length of monitoring term: in case of weekly, criteria for heavy rainfall are determined with empirical relationships based on monthly totals measured at stations. In some cases, historical daily records are too short to determine thresholds for weekly monitoring of heavy rainfall events. Some CPC products have been used to evaluate convective activities, but not systematically due to their insufficient spatial resolution. TRMM products were not picked up by the RCC, but GPM is

13

of interest due to its improved resolution, however interface for accessing products needs to be improved. The JRA reanalysis has a resolution of 0.625°. It was argued whether or not reanalyses were capable of correctly reflecting cloud and precipitation conditions, including extremes. The European Emergency Management Service is using reanalyses because they deliver a physically-consistent set of variables used in operational flood, drought, and forest fire operational services. 4.3 CMA/Beijing Climate Center Fengjin Xiao (not attending) provided slides on climate events monitoring and satellite application in China. The Centre is a recognized WMO RCC since 2009. In China, more than half of meteorological events are droughts, and 28% floods. The frequency and areas affected by droughts has increased in the past 60 years. Annual average of rain days has decreased by 13% and rainstorm day number has increased by 10%. Heat wave events have also increased. The Centre barely uses satellite data for monitoring extremes, but has identified a great demand. There is a severe lack of capable personnel and related technology. 4.4 Australian Bureau of Meteorology Yuriy Kuleshov provided an overview of climate statements issued by the AuBOM, and on the utilization by AuBOM of satellite data: AMVs derived from Himawari-8 are of key importance, and so are precipitation estimates, monthly NDVI maps and anomalies, and daily solar exposure. From a RA V perspective, small-island developing states (SIDS) are mostly vulnerable to Tropical Cyclones (TC). He quoted the examples of TC Tracy hitting Darwin in 1974, and TC Pam on Vanuatu in 2015, affecting 50% of the population. The 2012 Fiji floods consisted of two consecutive heavy rainfall events due to tropical depressions. The AuBOM international initiative on early warnings and preparedness to hydro-meteorological hazards aims to assist SIDS in Pacific Islands. Meteorological droughts affected Fiji, Tuvalu, and Samoa related to La Niña in 2011. The WMO RA V Pacific RCC network is under development – working arrangements and next steps in implementation and designation were discussed at a November 2016 meeting in Honolulu, USA. Except for Australia and Fiji, NMHS in the region do not use satellite-derived products. It was felt important to separate the monitoring of TCs from that of less intense systems that are often responsible for flooding events due to their slower movement. 4.5 BMKG Indonesia Riris Adriyanto briefly described how satellite observation is utilized by BMKG Indonesia for weather and climate monitoring. At present major satellite applications are on nowcasting and short-range weather forecasts, forest fire and smoke detection, volcanic ash monitoring. BMKG has been using Himawari-8 data received from the JMA HimawariCloud in operational weather forecasting, and GSMaP near-real-time data from JAXA-EORC has also been used for precipitation monitoring, weather analysis and forecast verification.

14

For climate-related phenomena, satellite-derived products have been also used for monitoring of MJO onset and propagation which has significant impacts on weather and climate in Indonesia. BMKG will proceed to a demonstration phase as the South-East Asia RCC Network in July 2017. This is in line with the BMKG plan to develop satellite-based products for climate-related phenomena such as drought monitoring. Due to the increasing importance of satellites as an Earth observation backbone, and recent significant improvements in new sensors onboard new generation satellites, it is expected that BMKG will be able to improve its satellite-based products to complement in-situ observations in monitoring climate extreme events. 4.6 Tanzania Meteorological Agency Ladislaus Chang’a provided an overview on the climate of Tanzania and main climate variability factors with common weather and climate extremes in the region, socio-economic impacts of climate extremes, and also challenges in utilization of the space-based weather and climate products. In Tanzania, common weather and climate extremes are heavy rainfall, floods and droughts, and high temperatures (heat waves) are becoming a problem. In the utilization of satellite derived products for monitoring weather and climate extremes, it is expected to enhance observation and monitoring of soil moisture, drought, heavy precipitation with enhanced capacities in processing and interpretation of products to generate tailor made climate information and products. It is also expected to enhance and ensure accessibility of those products by strengthening ground receiving capabilities.

5 CONTRIBUTION OF SATELLITE OPERATORS 5.1 JAXA/EORC Riko Oki introduced the GSMaP (Global Satellite Mapping of Precipitation) product developed by JAXA for GPM (latest v7 as of 17 Jan 2017). She showed an example of heavy rainfall over Thailand in Jan 2017 detected by the product in small river basin. A real-time version has been developed, using data with latency 0.5 hour and a forward extrapolation of 0.5 hour. Since Nov 2015, this product is available online; the new GSMaP RIKEN nowcast opens soon, using NRT (-4h) + 12 hour extrapolation. GSMaP has about 2600 users, more than half of which from outside Japan. She analysed the Lao PDR 2015 rainfall event to compare IMERG and GSMaP products. IMERG seems to have over-estimated rainfall, however GSMaP often under-estimates. Comparison of GSMaP and weather radar-based estimates shows that best correlation is obtained for 1.0deg spatially sampled GSMaP (for all temporal sampling), not at higher resolution. Assimilation of GPM/DPR data in the NWP system of JMA yields strong improvement in precipitation forecasts. Radar and satellite signals saturate at certain heavy rainfall rates; averaging over larger areas is needed to alleviate this problem. 5.2 CMA/NSMC Yang Jun provided an overview of the CMA satellite programme, both in low-Earth orbit (FY-3) and geostationary orbit (FY-2 and FY-4). The launch date of FY-3D has been postponed to September 2017: in its payload, it will have enhanced as well as new instruments. The FY-3RM planned for

15

after 2020 will have a dual-frequency precipitation radar on board, to provide 3D precipitation structure over ocean and land. Recently launched FY-4A instrumentation and baseline products, first image and first spectral signatures look promising. NSMC supports the Beijing national climate centre through reprocessing of CDRs, and the provision of operational (daily, weekly, monthly) products. He provided examples for monitoring of heavy precipitation, and fire maps. Several CDRs have been produced but still undergo evaluation. CMA agreed to include these in the ECV Inventory once quality checks have been completed. 5.3 NOAA/NESDIS Ralph Ferraro briefed on NESDIS activities in support of weather and climate extremes monitoring. He highlighted that GOES-16 (launched in November 2016) first imagery is now available and other products will be coming online soon; JPSS-1 satellite will be launched September 2017. He showed a list of precipitation, snow, soil moisture, water vapour and land surface products that are of potential interest to WMO RCCs. There are blended products on soil moisture (SMOPS) and water vapour. The ensemble tropical rainfall potential (eTRaP) forecasts probabilities of rainfall for land falling tropical systems based on accuracy and latency, using MW imagery. The HydroEstimator provides an IR-based rapid update technique to estimate rain rate. Lightning intensity (currently surface-based) may have predictive skill for tornado formation as well as vastly improve convective rain classification over traditional IR schemes. Direct readout MW data (with much lower data latency) is used to map snowfall intensity, especially over mountainous areas. Water vapour products can be used as a proxy for rainfall because they can track and monitor atmospheric rivers that are often associated with prolonged flooding events, and can help to estimate flooding potential. There are also emerging, new satellite-driven drought monitoring products. In addition, the National Center for Environmental Information (NCEI) manages the climate data record programme and has four precipitation CDRs; they produced comparisons of PERSIANN, CMORPH and GPCP long-term records and noted their similarities and differences, as well of their performance in extreme precipitation. Finally, he noted that NOAA/NESDIS/NCEI regional climate services are involved in RCC activities, in particular, in RA-V where satellites can help fill gaps, and to help monitoring oceans and coastal regions. 5.4 EUMETSAT Jörg Schulz presented on aspects of climate monitoring performed by EUMETSAT. He gave a view on very early geostationary imagery from 1966 and 1967. A year’s worth of Meteosat-1 data was found in old archives and needs to be analysed for being added to the climate record. Overlap of Metop-A, -B, and –C is likely and allows special insight into atmospheric sounding. Regional reanalyses (UERRA) do not make much use of high-resolution satellite data. EUMETSAT contributes to producing ECV records, both at headquarters and in the SAFs. Re-calibration of IR and WV channels over the entire Meteosat record is underway. The LST record is very stable over >25 years, potentially providing information for the analysis of heat waves. A recent EUMETSAT Climate Data User Feedback workshop shed light on the needs of

16

climate modellers; it was especially interesting to learn about the needs of decadal prediction models for observational records. A similar exercise could be done for the analysis of extremes. Jörg Schulz gave a short overview of the ECV Inventory of satellite-based climate data records currently under development by the CEOS-CGMS Working Group Climate. It currently has more than 750 entries. 6 REGIONAL CLIMATE MODEL FOR WEATHER AND CLIMATE MONITORING Dick Dee from the European Commission Copernicus Climate Change Service (C3S) hosted by ECMWF provided an overview of the Service and its vision. The Service takes advantage of the tools available for weather forecasting. Monthly updates of seasonal forecasts are available. Attribution of events to climate change has so far not been part of the Service since it is considered a topic of active research (in the EUCLEIA project, for example). Resolution is a major impediment for using reanalyses for monitoring extremes. Use of high-resolution observations over land is a particular challenge, since models generally do not accurately represent small-scale features over land. The C3S Sectorial Information System is to demonstrate the capability of the C3S Climate Data Store to meet the needs of some sectors, not to meet the needs of all users everywhere. High temporal resolution is also important to many users; updates every 30min or 10min would be useful especially with regard to extremes (currently output is hourly). Much progress was made to have trends represented in reanalyses, consistent with observations (where there are observations).

7 BREAK-OUT SESSIONS Key questions to be addressed by the break-out groups were:

Is operational monitoring of extremes from space possible?

Is the time ripe to introduce such monitoring in operational centres?

Do we have the building blocks available?

Can we get to an operational system?

What needs to be done next? It was noted that in terms of data provision, the necessary building blocks were in place; however on the part of intermediate users, they were mostly not in place in the climate domain and require attention by GFCS and other mechanisms. Especially WMO RCCs need more resources, better distribution, and better dialogue between satellite providers. 7.1 SESSION 1 This group consisted of: Ralph Ferraro (chair), Pingping Xie (rapporteur), Yuriy Kuleshov, Riris Adriyanto, Ladislaus Changa, Einar Bjorgo, Peter Salamon, Kiyotoshi Takahashi, Don Hinsman. Two overarching questions were assigned to this breakout group.

17

(1) To present use-cases for satellite-derived products focusing on monitoring extreme events with regard to ‘accumulated high precipitation’ and ‘drought’ (2) To understand and document the requirements from RCCs/climate centres for monitoring weather and climate extremes in operations, with a focus on ‘accumulated high precipitation’ and ‘drought’ In response to question (1), the group attempted to identify the satellite product baselines. It was found that several satellite operators (e.g., EUMETSAT, JAXA, NMSC, NOAA) routinely generate a variety of products on various time/space scales that are used, in some cases, by National Climate Centers and other affiliated agencies. Products such as precipitation rate/accumulation, water vapour, surface properties (vegetation parameters, soil moisture, surface temperature) are ‘baseline’ products generated by both low earth orbiting (LEO) and geostationary orbiting (GEO) satellites. These are typically level 2 (“L2”) products, useful for ‘weather’ extremes; L3 products (time and space averaged L2 data) are generated for utilization for ‘climate’ extremes. In some instances, there are fused data products, combining similar measurements from both LEO and GEO satellites, and in some cases, adding in-situ measurements where available. Examples of products that support monitoring and prediction of tropical cyclones, generated by NMSC and NOAA, are provided in the figures 7.1.1 below.

(a) (b) Figure 7.1.1: Operational applications based on satellite derived products in weather and climate events monitoring; (a) typhoon monitoring in National Satellite Meteorological Center (NSMC), CMA, (b) Forecast of 24-hour rainfall potential for tropical systems by Ensemble Tropical Rainfall Potential (eTRaP), NOAA/NESDIS There is a wide range of the uses of such products by national, RCCs and other supporting groups such as GEO, GFCS and UNOSAT. In some cases, satellite products are used in conjunction with in-situ, whereas some RCCs are not yet exploiting satellite products, particularly those centers that have limited computer/communication capabilities. Additionally, these centers also are not fully aware of which products are available or which ones to use. Finally, for those centers using satellite products, there are instances where they work well (e.g. using NOAA OLR for monitoring Convections), others where they do not (e.g. relying on CLIMAT for monitoring Precipitation and Temperature), are provided in the figures 7.1.2 below.

18

Figure 7.1.2: Summary of the 2016 Asian Summer Monsoon; Tokyo Climate Center (TCC)

News, No.46, Autumn 2016, Japan Meteorological Agency

(http://ds.data.jma.go.jp/tcc/tcc/news/tccnews46.pdf)

In response to question (2), the group discussed what the requirements were for monitoring weather and climate extremes. In general, the RCCs need information on drought, precipitation, heat/cold wave, and storms. There are two types of droughts: agricultural (growing season) and hydrological (steam flow, ground water and long term storage). Monitoring them require entirely different measurements and products. Agricultural drought requires temperature, vegetation parameters (e.g., NDVI, LAI, etc.), soil moisture, and precipitation. The representatives from the RCCs noted that knowing the onset and duration of a drought are critical pieces of information. Hydrological drought requires vastly different information and its monitoring is especially critical in regions with pronounced wet seasons. High spatial resolution measurements (less than 500 m) of river/lake/reservoir levels are needed. Additionally, measurements of ground water are desirable.

19

Some of these are only available from research satellite platforms/sensors. Additionally, snow water equivalent of snowpack is also important. It was noted that these types of measurements have strong seasonal and regional dependencies. The figure 7.1.3 below provides an example of a useful surface type product derived from a low-earth orbiting satellite, S-NPP, and the VIIRS sensor.

Figure 7.1.3: NOAA Surface Type Environmental Data Record (ST-EDR)

(https://viirsland.gsfc.nasa.gov/NOAAprod.html)

Precipitation extremes fall into three broad categories – local/flash floods, “synoptic scale”/longer duration events, and seasonal/inter-annual (i.e., ENSO, MJO) events. In many cases, they are all “connected” as certain seasonal phenomenon lead to increased likelihood of flash flooding or more synoptic scale storms. The product needs are different for each of these. For example, geostationary orbiting measurements are most critical for the short-lived, very local-scaled flash floods whereas merged satellite products such as GSMaP (see figure 7.1.4 below), IMERG, and CMORPH are particularly useful for the latter two types. Additionally, CDRs (including ICDRs) are useful for the third type.

Figure 7.1.4: Total precipitation accumulation from GSMaP [08 Aug. –25 Aug. 2016]

(http://sharaku.eorc.jaxa.jp/GSMaP/)

20

Temperature extremes from prolonged heat and cold waves can have tremendous impact on society and agriculture. Their impact is seasonally and regionally driven. It was also noted that low level moisture can also increase the hardship in heat waves. Finally, the group discussed “Storms and High Winds”. Two types exist and are more or less likely under certain seasonal climate variations. In the larger scale, high winds associated from tropical cyclone and mid-latitude systems impact coastal regions and offshore waters. On the smaller scale, very localized, downburst or tornadic storm damage should be considered, although these were determined to be more of a local weather forecast concern than one for RCCs.

21

The group then developed RCCs product requirements, which are summarized in the table below.

Phenomenon Space-time

grid

Areal coverage /

resolution

Temporal

coverage /

resolution

Update

frequency

Latency Confidence in index

(accuracy) requirement

Comments

Drought 0.5deg

7-10 day

global land 1981-present consistent with

time resolution

up to 2

days

Qualitative index Precipitation, T2m, ETP, soil

moisture, water levels,

groundwater, river runoff,

vegetation index

Large-scale

flooding

1-10-km

3-hly

basin – global 10-yr or longer consistent with

time resolution

<= Time

resolution

% error

Probability of Exceedance

Heat/cold wave 10-km

daily

regional 1981-present daily <= 1day In deg C or % error Daily Max/Min/mean T and

moisture, wind speed

Wind storm /

large scale

10-km

hly

regional / global 10-yr or longer consistent with

time resolution

<=Time

resolution

% error Surface wind speed / direction,

lightning, convective index

Fig 7.1.5: Space-time domain of weather, climatic, and flooding events.(Katie Hirschboeck, The University of Arizona) (http://www.southwestclimatechange.org/impacts/water/floods#references)

22

The group was charged by the meeting conveners to consider other aspects related to the use of satellite products. These are presented below:

• Does the WMO definition of extremes served the needs of the RCCs? - The answer is yes.

• How do you access and use satellite products and what challenges do you face? - It varies amongst the different RCC’s present. Some centers can access a considerable amount of data from various satellite operators, others are very limited due to slow communications and computing facilities

• How useful are NWP reanalysis and satellite derived ICDRs? - NWP reanalysis will be extremely useful for developing baselines for common meteorological variables like temperature and humidity. However, the current generation reanalysis will not be useful for precipitation because of model inadequacies and poor spatial resolution. It was noted by data providers that the reanalysis data will be useful for improving satellite products; some precipitation products today use NWP temperature fields as part of their retrieval. The ICDRs will be good for continued monitoring; having a lower latency than a CDR should outweigh the slight loss in accuracy (it was noted that today’s era of satellites are much better calibrated from inception so current ICDRs should be relatively well calibrated). However, it was also pointed out that we need to see through actual practice by the RCCs and national forecast centers what error tolerance is acceptable.

• Are we ready to move forward for operational use of satellite products at RCCs? - The group felt that yes, we are. There are several pieces in place to move forward, including:

– Space systems/measurements; there are more than enough in place to address the problem of monitoring extremes. Of course improvements in the future will be needed, such as higher spatial resolution, more routine measurements from certain payloads, etc.

– Ground systems to generate products are mature by many operational and research satellite centres. There are a few areas of concern such are the proper synchronization between satellite operators and RCC end user during algorithm updates and subsequent reprocessing and are the products generated meeting the RCC requirements (as specified in the preceding table).

– Data delivery/pipelines are generally in place and reliable

• We noted some things that are lacking: – Not all RCCs are ready yet, but we can certainly define some pilot projects for next

1-2 years to get started – We do not fully understand the needs of the RCCs – a clearer definition of

requirements is needed – Education and training – Centralized data portal and product lines (be good to have consistency on what we

provide and use)

RECOMMENDATIONS AND ACTIONS

RECOMMENDATION 1: To CGMS (and CEOS) – There is a continuing need for improved satellite payloads. We recommend that agencies continue to strive for improved sensors as baseline missions that consider the following attributes: improved spatial resolutions for all sensors; expanded baseline payloads as new research technologies are proven (e.g., space based radars for precipitation; high resolution SAR/scatterometer for surface information, etc.)

RECOMMENDATION 2: To CGMS (and CEOS) – Support the R&D needed to exploit emerging technologies (e.g., lightning, multi-sensor and in-situ data fusion, etc.) in order to improve precipitation retrieval in deficient precipitation regimes, including orographic/”warm top”

23

precipitation and flash floods/short duration events that rely on geostationary orbiting measurements (through use of lightning observations); improve the representation of heavy precipitation, reducing regional biases, in particular, those associated with tropical systems; exploitation of reanalysis data sets. RECOMMENDATION 3: To WMO Secretariat – Develop a survey among all WMO RCCs and their end users to determine their product needs for monitoring weather and climate extreme events. Survey should include product attributes, uncertainty needs, etc. RECOMMENDATION 4: To WMO Secretariat – Define with the WMO RCCs, NMHSs and relevant institutes in East Asia/Western Pacific region to set up an initial demonstration project (Space-based Weather and Climate Extremes Monitoring Demonstration Project (SEMDP)). Future demonstration projects will be extended to other WMO regions based on the experience of this initial project. RECOMMENDATION 5: To WMO Secretariat – Establish an ad-hoc expert team, consisting of Satellite Operators, R&D Space Agencies, WMO RCCs, NMHSs and relevant institutes, for drafting an implementation plan for SEMDP in the East Asia/Western Pacific region; the Project should demonstrate the use of existing and newly developed satellite- derived products in quasi-real time operations; products should consist of time series of measurements specific to the regional and national levels, along with related in-situ and/or model reanalysis data, and incorporating relevant research RECOMMENDATION 6: To WMO Secretariat – Develop a first draft of a concept paper outlining a structure of an operational Space-based Weather and Climate Extremes Monitoring system (SWCEM) to be discussed at the upcoming meeting of CGMS and CEOS. RECOMMENDATION 7: To WMO Secretariat – Enhance the awareness on the potential and application of space-based weather and climate products. In spite of the progress and potential of space-based weather and climate products, their utility is low at the WMO RCCs and NMHSs in developing countries, particularly in Africa. Contributing to this are the lack of awareness about the products, their reliability, and the lack of technical capacity in accessing and processing the products.

RECOMMENDATION 8: To WMO Secretariat – Ensure the proper user-provider feedback mechanism, a proper platform for the RCCs and NMHS’s to provide continuous feedback on the effectiveness and challenges in the application of these products needs to be developed. 7.2 SESSION 2

This group consisted of: Rainer Hollmann (chair), Jörg Schulz, Mark Dowell, Riko Oki, Tillmann Mohr, Yang Jun, Ali Behrangi, Simon Eggleston, Erica Allis, Andre Obregon Two overarching questions were assigned to this breakout group.

(1) To identify lessons learned in the formulation of weather and climate extremes-related

requirements and satellite-specific responses, and to formulate a preliminary response by RCCs/climate centres to the identified requirements

(2) To assess the feasibility of a demonstration of space-based weather and climate extremes monitoring in operations, for strengthening the capacity of NMHSs.

Figure 7.2.1 below, taken from Satellite Climate Data Records: Development, Applications, and Societal Benefits; Wenza Yang et. al.; Remote Sens. 2016, 8, 331; doi:10.3390/rs8040331, as well as the Architecture for Climate from Space show the different production pathways for climate data records. In the use of satellite data to support monitoring of climate extremes, the group discussed

24

the components of the observing system needed to perform it. Currently often anomalies are derived from operational NRT (Near Real Time) products which may not fulfill climate requirements. Because of the need for quantitative anomalies, the ambition for a regularly updated climate data record (ICDR) shall be its maximum consistency with climatology. This can be ensured in that way that for an ICDR a processing system as close as it has been used in the generation of the related TCDR has been used. The group assumed that a 5-day latency in the provision of ICDR products would fulfill most application areas (e.g., including GSMap, IMERG). The provision of ICDR is performed by a few satellite operators (e.g. EUMETSAT through its CM SAF), but a truly global coverage of ICDR products is so far not achieved (e.g. precipitation is usually not available for high latitudes). Many users will need satellite-based products with short latency (on the order of hours to days), for monitoring short-term climate variability, and weather and climate extremes. CDR and ICDR products are desirable to relate the latest measurements to the climatological reference. A clear difference between a NRT and an ICDR product is that the first one takes benefit from the most recent scientific developments, capabilities and improvements of the (satellite) instruments. An ICDR is usually limited to the common instrumental capabilities, supporting a more homogenous time-series. The group realized that more research efforts are needed to establish the scientific background and to showcase the relative merit of consistency of products with the climate record needs.

Figure 7.2.1: Schematic diagram of production pathway for various climate data record

products.

Finally, the group noted, that the definition of a TCDR is being done in the “sustained architecture of climate” whereas a common and accepted definition of an ICDR is still needed to provide best practices to RCCs in setting up operations on extreme monitoring.

The definition used by the CM SAF could be adopted which defines an Interim Climate Data Record as a CDR regularly updated with an algorithm / system having maximum consistency to TCDR generation algorithm / system. The update cycle depends on the user needs for climate extremes and might range from pentad to monthly.

25

The group was of the opinion that in terms of governance an ICDR data generation is best done by the producer of the climatology (FCDR and TCDR) or in very close connection to them to enable and ensure an utmost consistency.

Furthermore, the group noted that he maturity matrix is largely applicable to this with some

changes needed in uncertainty characterization (see summary of further discussion below). However, the group noted that more feedback from the community is needed to trigger feedback.

RECOMMENDATION 9: The CEOS/CGMS Working Group on Climate to adopt the above proposed definition of an ICDR as part of the Climate Architecture.

RECOMMENDATION 10: The CEOS/CGMS Working Group on Climate to confirm as best practice that the producer of the climatology (FCDR and TCDR) is best positioned to be responsible for an ICDR, in order to enable and maximize consistency. RECOMMENDATION 11: Data providers to provide at least two kinds of datasets for different applications, one being consistent with the longest available CDR, and the other one taking into account the latest developments/capabilities of instruments. The group then discussed about climate indices from ETCCDI and whether they can be used in the assessment of extremes. It was noted that indices (based on those adopted and developed by ETCCDI) are based on percentile approach applied on the statistical distribution of precipitation and temperature. A more general definition of extremes and characterization of extreme weather and climate events is currently subject of work by the Commission for Climatology Task Team on the Definition of Extreme Weather and Climate Events (TT-DEWCE)7. So if there is a short record, there are thresholds beyond which things are anyway extreme. The mixture may introduce uncertainty in the indices, e.g., different products differ in indices. They converge only at some aggregated space-time resolution. Thus the group realized that the impact areas of the RCCs to be addressed are regionally dependent (they do not work everywhere) and identified a clear need for regional indicators. In addition, taking into account limited feedback from RCCs, the group took note that there is additional scope for indices taking into account impacts on ecosystems (e.g., drought onset indicator is a real user driven indicator. There seem to specific timings in the agriculture calendar planning (1st April seems important) that may not match with the precipitation. In conclusion, indicators on timing of precipitation are important as well. Impacts are stronger when they are overlaid by other conditions. There is need of integrating several data sources including social indicators such as health indicators, malaria etc. Some indicators may point to cumulative effects of conditions. To support Operational Space-based Weather and Climate Extremes Monitoring by the RCCs, further needs have been identified: • To properly define and characterize the extreme events with their various aspects: location,

time, severity and extent to support operational monitoring, and provide guidance on this • There is a clear need for additional data compared to the traditional climatological data (near

surface precipitation and accumulated precipitation) • From satellite this could be precipitation, soil moisture, vegetation status, surface temperature • Existing indicators do not always take into account ecosystem impacts • RCCs should select relevant sets of indicators for their regions. • A central database of regions and their indicators could be helpful to implement this.

7 http://www.wmo.int/pages/prog/wcp/ccl/opace/opace2/TT-DEWCE-2-2.php

26

RECOMMENDATION 12: Given the known differences between ground station and satellite observation, it is recommended that ETCCDI and the WMO Commission for Climatology relevant groups8 to consider, identify, and/or develop data requirements to help in developing jointly with space community indices and analysis methods for assessing and monitoring extreme weather and climate events based on space type observations, taking into account the needs of intermediate users (e.g. RCCs). RECOMMENDATION 13: The ETCCDI team to develop guidance on the regionalization of (satellite-based) climate indices together with RCCs, and communicate this to the RCCs. The group then discussed about uncertainty information and their provision to users of the RCCs. The group sees that uncertainties are an important element to allow RCCs and other Services for making selections of offered data sets. Requirement tables may contain a box for type of uncertainty information and tolerable uncertainty. The group recognized the need to have tailoring of uncertainty messages (see Otto et al. (2016)9). In summary, uncertainty is • Absolutely needed for data assimilation or model implementation. • Needed in communication to users, rather as confidence or probability of something is likely to

happen. • Needed geographically distributed and static per season but ideally it should be fully dynamic. • What type of uncertainty information should be provided? Providing very detailed information

vs. spread of uncertainty of different products. Following this, the group sees the RCC as an important mediator in the chain. The RCCs are seen as “translator” of the full available information coming from the CDR generator to an more “easy to understand” for the RCC user. RECOMMENDATION 14: Organize work with RCCs on usage of and needs for uncertainty information, for example through the Joint CCl-CBS expert team on RCCs. In addition, RCCs to consider how to best communicate product uncertainty to their user groups. It is recommended to implement this by means of RCC co-location meetings including relevant experts. MEETING URL: http://www.wmo.int/pages/prog/sat/meetings/workshop_on_SWCEM/WorkshoponSWCEM.html

_______________________

8 http://www.wmo.int/pages/prog/wcp/ccl/opace/opace2-ccl16.php

9 Otto et al. (2016) : Uncertainty: Lessons Learned for Climate Services. Bulletin of the American

Meteorological Society, December 2016, DOI: http://dx.doi.org/10.1175/BAMS-D-16-0173.1

27

APPENDIX A: LIST OF PARTICIPANTS

Category Name Organization

Satellite Operator

Yang Jun CMA/NSMC

Joerg Schulz EUMETSAT

Riko OKI JAXA/EORC

Ralph Ferraro NOAA/NESDIS

Regional Climate Centre /

NMHS

Yuriy Kuleshov AuBoM

Riris Adriyanto BMKG

Fengjing XIAO (not

attending)

CMA/NCC

Kiyotoshi TAKAHASHI JMA/CPD

Pingping Xie NOAA/NWS/CPC

Ladislaus Chang’a Tanzania Meteorological

Agency

International Science WG

CM-SAF Rainer Hollmann DWD

GEWEX Joerg Schulz EUMETSAT

WCRP/ETCCDI Ali Behrangi NASA JPL

SCOPE-CM Jeff Privette (remote) NOAA/NCEI

IPWG Ralph Ferraro NOAA/NESDIS

Space Programme fathers

Consultant for WMO SG Tillmann Mohr

Donald Hinsman

UN and Intergovernmental

Organizations

Mark Dowell European Commission Joint

Research Centre (EC/JRC)

Peter Salamon European Commission Joint

Research Centre (EC/JRC)

Copernicus Emergency

Management Service

Dick Dee ECMWF (European Copernicus

Climate Change Service

(C3S))

Einar Bjorgo UNITAR/UNOSAT

WMO Secretariat

Space Programme Toshiyuki Kurino

Space Programme Stephan Bojinski

GCOS Carolin Richter

GCOS Simon Eggleston

WCRP Boram Lee

GFCS Filipe Lucio

GFCS Erica Allis

CLW/DMR Omar Baddour

CLW/Agricultural Met Robert Stefanski

CLW/Agricultural Met José Camacho

GEO

Barbara Ryan

Andre Obregon

Vanessa Aellen

Osamu Ochiai

28

APPENDIX B: WORKSHOP RECOMMENDATIONS

RECOMMENDATIONS

RECOMMENDATION 1: To continue to strive for improved sensors as baseline missions that consider the following attributes: improved spatial resolutions for all sensors; expanded baseline payloads as new research technologies are proven (e.g., space based radars for precipitation; high resolution SAR/scatterometer for surface information, etc.).

CGMS (and CEOS)

RECOMMENDATION 2: To support the R&D needed to exploit emerging technologies (e.g., lightning, multi-sensor and in-situ data fusion, etc.) in order to improve precipitation retrieval in deficient precipitation regimes, including orographic/”warm top” precipitation and flash floods/short duration events that rely on geostationary orbiting measurements (through use of lightning observations); improve the representation of heavy precipitation, reducing regional biases, in particular, those associated with tropical systems; exploitation of reanalysis data sets.

CGMS (and CEOS)

RECOMMENDATION 3: To develop a survey to send to all RCCs and their end users to determine their product needs for monitoring weather and climate extreme events. Survey should include product attributes, uncertainty needs, etc.

WMO Secretariat

RECOMMENDATION 4: Define with the WMO RCCs, NMHSs and relevant institutes in East Asia/Western Pacific region to set up an initial demonstration project (Space-based Weather and Climate Extremes Monitoring Demonstration Project (SEMDP)). Future demonstration projects will be extended to other WMO regions based on the experience of this initial project.

WMO Secretariat

RECOMMENDATION 5: To establish an ad-hoc expert team, consisting of Satellite Operators, R&D Space Agencies, WMO RCCs, NMHSs and relevant institutes, for drafting an implementation plan for SEMDP in East Asia/Western Pacific region; the Project should demonstrate the use of existing and newly developed satellite- derived products in quasi-real time operations; products should consist of time series of measurements specific to the regional and national levels, along with related in-situ and/or model reanalysis data, and incorporating relevant research.

WMO Secretariat

RECOMMENDATION 6: To develop a first draft of a concept paper outlining a structure of an operational SWCEM to be discussed at the upcoming meeting of CGMS and CEOS.

WMO Secretariat

RECOMMENDATION 7: To enhance the awareness on the potential and application of space-based weather and climate products. In-spite of the progress and potential of space-based weather and climate products,

WMO Secretariat

29

their utility is low at the NMHS and RCC in developing countries, particularly in Africa. Contributing to this are the lack of awareness about the products, their reliability, and the lack of technical capacity in accessing and processing the products.

RECOMMENDATION 8: To ensure the proper user-provider feedback mechanism, a proper platform for the RCC’s and NMHS’s to provide continuous feedback on the effectiveness and challenges in the application of these products needs to be developed.

WMO Secretariat

RECOMMENDATION 9: To adopt the above proposed definition of an ICDR as part of the Climate Architecture.

The CEOS/CGMS WG on Climate

RECOMMENDATION 10: To confirm as best practice that the producer of the climatology (FCDR and TCDR) is best positioned to be responsible for an ICDR, in order to enable and maximize consistency.

The CEOS/CGMS WG on Climate

RECOMMENDATION 11: To provide at least two kinds of datasets for different applications, one being consistent with the longest available CDR, and the other one taking into account the latest developments/capabilities of instruments.

Satellite Derived Products Providers

RECOMMENDATION 12: Given the known differences between ground station and satellite observation, to consider, identify, and/or develop data requirements to help in developing jointly with space community indices and analysis methods for assessing and monitoring extreme weather and climate events based on space type observations, taking into account the needs of intermediate users (e.g. RCCs).

CCl/WCRP/JCOMM ETCCDI and the

WMO Commission for Climatology

relevant groups10

RECOMMENDATION 13: To develop guidance on the regionalization of (satellite-based) climate indices together with RCC’s, and communicate this to the RCCs.

CCl/WCRP/JCOMM ETCCDI Team

RECOMMENDATION 14: To organize work with RCCs on usage of and needs for uncertainty information, for example through the Joint CCl-CBS expert team on RCCs. In addition, RCCs to consider how to best communicate product uncertainty to their user groups. It is recommended to implement this by means of RCC co-location meetings including relevant experts.

WMO Secretariat

10

http://www.wmo.int/pages/prog/wcp/ccl/opace/opace2-ccl16.php

30

APPENDIX C: Acronyms and Abbreviations CDR Climate Data Record CEOS Committee on Earth Observation Satellites CGMS Coordination Group for Meteorological Satellites CLiC Climate and Cryosphere Project (WCRP Core-Project) CLIVAR Climate and Ocean – Variability, Predictability, and Change (WCRP Core-Project) CORDEX Coordinated Regional Climate Downscaling Experiment (WCRP) CLIMDEX Datasets for Indices of Climate Extremes CMA China Meteorology Administration CMAP CPC Merged Analysis of Precipitation CMIP WCRP Coupled Model Intercomparison Project CMORPH CPC Morphing technique CPC Climate Prediction Center ECV Essential Climate Variable EDR Environmental data Record ENSO El Nino Southern Oscillation ETCCDI CCl/WCRP/JCOMM Expert Team on Climate Change Detection and Indices ETP Evapotranspiration eTRap Ensemble Tropical Rainfall Potential EUMETSAT European organisation for the exploitation of METorological SATellites GASS GEWEX Global Atmospheric System Studies Panel (WCRP) GC-Extremes WCRP Grand Challenge on Understanding and Predicting Weather and Climate

Extremes GCOS Global Climate Observing System GEWEX Global Energy and Water Cycle Exchanges (WCRP Core-Project) GEO Geostationary Satellite GEO Group on Earth Observation GDAP GEWEX Data and Assessment Panel (WCRP) GFCS Global Framework for Climate Services GHP GEWEX Hydroclimatology Panel (WCRP) GLASS GEWEX Global Land/Atmosphere System Study (WCRP) GPCC Global Precipitation Climatology Centre GPM Global Precipitation Measurement mission (NASA) GSMaP Global Satellite Mapping of Precipitation GW Ground Water ICDR Interim Climate Data Record IMERG Integrated Multi-satellite retrievals for GPM IPCC Intergovernmental Panel on Climate Change IPWG International Precipitation Working Group JAXA Japan Aerospace Ecporation Agency JMA Japan Meteorological Agency LAI Leaf Area Index LEO Low Earth Orbit MJO Madden-Julian Oscillation NDVI Normalized Difference vegetation Index NESDIS National Environmental Satellite, Data, and Information Service NMSC National Meteorological Satellite Center NMHS National Meteorological and Hydrological Service NOAA National Oceanic and Atmospheric Administration OLR Outgoing Longwave Radiation PDF Probability Density Function RCC Regional Climate Center SDG Sustainable Development Goal SM Soil Moisture

31

SMOPS Soil Mositure Operational Product Systems (NESDIS) S-NPP Suomi National Polar-orbiting Partnership SPARC Stratospheric Processes and their Role in Climate (WCRP Core-Project) SST Sea Surface Temperature ST Surface Type TCC Tokyo Climate Center TRMM Tropical Rainfall Measurement Mission UNOSAT Operational Satellite Applications Program of the UN Institute for Training and

Research (UNITAR) WCRP World Climate Research Programme WMO World Meteorological Organization