# Workshop: Observational campaigns for better weather forecasts

Europe/London
ECMWF

#### ECMWF

Description

This workshop aimed to strengthen the links between observational campaigns and Numerical Weather Prediction (NWP) to (1) help observational campaigns optimise their use of forecasts and knowledge of forecast errors and (2) help NWP system development make better use of observational campaign data. Forecast errors in NWP, such as in the ECMWF forecast system, often motivate field campaigns, and forecasts are used in many ways during these campaigns. This workshop invited experts involved in recent or future campaigns that have made or will make use of ECMWF data to share their experiences and provide feedback on their use of ECMWF forecasts. The process of campaign planning, diagnostics and improved understanding of model errors, and ultimately the improvement of NWP systems will be key themes of the workshop.

Registration
Workshop: Observational campaigns for better weather forecasts
Events team
• Monday, 10 June
• 12:30 13:00
Registration and coffee 30m
• 13:00 15:30
From models to observations and back - Chair: Linus Magnusson
• 13:00
Opening and welcome 15m
Speaker: Florian Pappenberger (ECMWF)
• 13:15
How ECMWF supports field campaigns 15m
Speaker: Emma Pidduck (ECMWF)
• 13:30
From Models to Data 30m

Often it is thought that observations exist to serve models. A commonplace idea in the atmospheric and climate sciences is that observations exist to 'verify', 'evaluate' or 'improve' models. In this talk I instead focus on how models can be used to guide observations. By identifying specific processes underlying a particular type of modeled behavior, models can instead be used to craft hypotheses that observations can be used to test. This is an old-fashioned an familiar idea for much of the physical sciences, but poorly practiced in atmospheric and climate science. Examples will be provided of where this has been successful, and why it might be useful to think of models as tools for guiding and sharpening observations, rather than the other way around.

Speaker: Bjorn Stevens (Max Planck Institute for Meteorology)
• 14:00
Airborne active remote-sensing observations of the extratropical troposphere and lower stratosphere with a special focus on the NAWDEX field experiment 30m

In the past decade, the German Aerospace Center was involved in a series of observational campaigns with various objectives, e.g. on atmospheric dynamics, cloud microphysics or atmospheric chemistry. This presentation gives an overview of active remote-sensing lidar and radar profile observations of winds, humidity, ozone, aerosols and clouds; i.e. parameters of potential relevance for numerical weather prediction (NWP).

The focus of the talk will be set on the North Atlantic Waveguide and Downstream impact EXperiment (NAWDEX), a multi-aircraft field campaign that was conducted over the North Atlantic Ocean in autumn 2016. NAWDEX was the first field experiment with synergistic airborne and ground-based observations from the entrance region to the exit region of the storm track, and was undertaken to investigate the role of diabatic processes in altering jet stream disturbances, their development, and their effects on high impact weather (HIW) downstream. We will illustrate how ECMWF deterministic and ensemble forecasting products were used to define the scientific goals, the experimental design and flight planning.

NAWDEX provides an excellent opportunity to study forecast errors as the campaign period contained episodes of reduced predictability, indicating that uncertainties originating in the estimated atmospheric state and model formulation grew rapidly. Weather features expected to be associated with forecast errors were extensively probed by accurate high resolution cross sections. As an example of systematic meteorological analysis errors, a study of the representation of jet stream winds during the NAWDEX period will be presented. A comprehensive set of wind profile observations across the tropopause from airborne lidar, dropsondes and a ground-based wind profiler was compared with analyses of ECMWF’s IFS and MetOffice’s Unified Model. The results look pretty similar for both models and revealed increased uncertainty of winds and wind speed gradients at and directly above the tropopause, especially in situations of elevated tropopauses downstream of mid-latitude cyclones. Short-term forecasts showed radidly growing errors in this region with an underestimation of wind speeds.

Additional to this study on winds, we will discuss previous results on the representation of lower tropospheric humidity together with plans for future work on validation of NWP models. Based on the described airborne observations, we aim to identify areas with potential for improved collaboration between the NAWDEX community and ECMWF.

Speaker: Andreas Schäfler (Deutsches Zentrum für Luft- und Raumfahrt)
• 14:30
Experiences from Arctic field campaigns: Weather forecasting and evaluations, and model comparisons 30m

Over the last decades the Swedish icebreaker Oden has had atmospheric research mission to the central Arctic four summers; 2001, 2008, 2014 & 2018. The data brought back from these expeditions provides invaluable detail on conditions and processes that can be used to test models and model formulations. But the field campaigns also rely on weather forecasts for the operations. One very special aspect of weather forecasting for central Arctic expeditions is the very limited bandwidth. The forecast material available to the ship’s meteorologist is typically very limited, a few forecast maps, some of the expeditions own observations and real-time satellite imagery from polar orbiters.

In this presentations we will explore the data gathered during such expeditions and how it has been used for model testing. We will also review what forecast material that has been available, with a focus on the latest endeavor: Arctic Ocean 2018 expedition (AO18). This expedition took place during August and September 2018, during the tail end of the Year of Polar Prediction (YOPP) Arctic Summer Special Observing Period (SOP). For AO18, YOPP in collaboration with the EU HORIZON2020 project APPLICATE provided a set of specialized forecast for AO18 based on the ECMWF IFS model; forecast were also evaluated in near-real time and some of the results will reviewed.

Speaker: Michael Tjernström (Stockholm University)
• 15:00
Synergetic use of field campaign observations and detailed simulations to improve NWP models at Météo-France 30m

Météo-France/CNRM has a long experience and know-how in designing field experiments for developing and improving physical parameterizations for atmosphere, continental surfaces and ocean coupled models. The most outstanding field campaigns of the last decades include those coordinating in CAPITOUL on urban boundary layer in 2004-2005 (Masson et al, 2008), in AMMA on African Monsoon in 2006 (Redelsperger et al, 2006), in HyMeX on Mediterranean heavy precipitation and flash-floods in 2012 (Ducrocq et al, 2014) and on dense water formation induced by strong regional winds in 2013 (Estournel et al, 2016) and, for the near future ones, SOFOG3D on fog in winter 2019-2020 and HyMeX-LIAISE on the human imprint on atmosphere-land surface Interactions over semiarid regions in spring and summer 2020.
The aims of the presentation is to describe the methodologies that are developed to make use of these field campaign observations for improving the physical parameterizations (e.g. microphysics, turbulence, land surface schemes, urban surface models, air-sea fluxes,…) of numerical models. The methodologies often include a combine use of field campaign observations and of detailed model simulations, such as Large Eddy Simulations.

Ducrocq, V., et al, 2014: HyMeX-SOP1, the field campaign dedicated to heavy precipitation and flash flooding in the northwestern Mediterranean. Bulletin of the American Meteorological Society, 95, 1083-1100.

Estournel, C., et al., 2016: HyMeX-SOP2: the field campaign dedicated to dense water formation in the northwestern Mediterranean. Oceanography, 29, 196-206.

Masson, V., et al.,2008: The Canopy and Aerosol Particles Interactions in TOulouse Urban Layer (CAPITOUL) experiment, Meteorol. Atmos. Phys., 102, 135–157.

Redelsperger, J., C.D. Thorncroft, A. Diedhiou, T. Lebel, D.J. Parker, and J. Polcher, 2006: African Monsoon Multidisciplinary Analysis: An International Research Project and Field Campaign. Bull. Amer. Meteor. Soc., 87, 1739–1746.

Speaker: Véronique Ducrocq (Météo-France/CNRM)
• 15:30 16:00
Coffee break 30m
• 16:00 17:30
From models to observations and back - Chair: Mohamed Dahoui
• 16:00
Application of Ensemble-based Sensitivity to ECMWF Ensemble Forecasts in Field Campaigns 30m

Although the most popular application of ensemble forecasts is the mean and forecast standard deviation, there is substantial information within the higher moment statistics of these datasets that can be used to evaluate the dynamics and predictability of dynamical systems. In addition, these ensemble-based sensitivity methods can be used to identify locations where additional observations could change and/or reduce the variance in a forecast metric of interest, such as tropical cyclone (TC) position or rainfall. One of the advantages of this approach is that it is a computationally inexpensive post-processing application that can be quickly produced given ensemble forecast output. The goal of this talk is to provide examples of how ensemble-based sensitivity applied to ECMWF ensemble forecasts has been used in past field programs. During 2017 and 2018, experimental ensemble-based sensitivity has been used to guide where to deploy aircraft dropwindsonde locations that are meant to improve TC track forecasts. In 2019, this technique will be used to evaluate the forecast uncertainty and target locations for landfalling atmospheric rivers along the west coast of North America.

Speaker: Ryan Torn (University at Albany, SUNY)
• 16:30
Observations to operations in the Met Office Unified Model 30m

Observation-model difference drive the development of parametrizations. Here we will show how the combination of aircraft, satellite and ground based observations as well as theoretical and seamless modelling has led to model improvements in the Met Office Unified Model. Examples to be discussed will include i) the Southern Ocean sea surface temperature bias that draws on a rich combination of field work, case studies and theoretical underpinning, ii) the global model representation of light rain that tries to capture the broad range of the rain drop distribution for an operational single moment representation based on extensive aircraft observations, iii) a multi modal pdf approach for cloud fraction tested against ARM site data and iv) the impact of an improved land surface representation from airborne operations.

Speaker: Paul Field (Met Office, Univ. of Leeds)
• 17:00
How ARM/ASR observations have contributed to ECMWF model development 30m
Speaker: Irina Sandu (ECMWF)
• 17:30 18:00
Operational briefing 30m Weather Room ()

### Weather Room

Speakers: Jeremy Dehart (USAF Reserve Hurricane Hunters), Ryan Rickert (53 WRS Hurricane Hunters)
• 18:00 19:30
Drinks reception and poster session Weather Room ()

### Weather Room

• Tuesday, 11 June
• 08:55 10:30
Diagnostics or making use of the data - Chair: Mark Rodwell
• 09:00
Is it time for interactivity and 3D? New approaches to analysing NWP data for observational campaigns using 3D and ensemble visualization 30m

Visualization is an important and ubiquitous tool in the daily work of atmospheric researchers to analyse data from simulations and observations, and field campaigns are no exception. Visualization techniques are applied during flight planning to analyse Numerical Weather Prediction (NWP) output to perform weather forecasting under campaign-specific requirements, for example, to predict the occurrence of warm conveyor belts or specific chemical species. Also, after a campaign has been conducted, analysis of collected cases heavily relies on visualization.

Visualization research has made much progress in recent years, in particular with respect to techniques for ensemble data, interactivity, 3D depiction, and feature-detection. Transfer of new techniques into the atmospheric sciences, however, is slow. This talk will present recent developments of interest to flight planning and case analysis, in particular focusing on the “Met.3D” project (https://met3d.wavestoweather.de), our effort to make novel 3D ensemble visualization techniques accessible to atmospheric researchers. We will discuss experiences gained during the 2016 North Atlantic Waveguide and Downstream Impact Experiment (NAWDEX), during which interactive 3D ensemble visualization was applied to analyse ECMWF forecasts for flight planning. The NAWDEX cases also subsequently received strong focus of research conducted on facilitating visual analysis of synoptic-scale atmospheric features including jet-stream core lines and fronts in midlatitude cyclones, and on interactive visual analysis of similarity and sensitivity within ensemble predictions. We will discuss the benefit of these new techniques for analysing campaign cases, and point out the potential of 3D interactive visual analysis as a diagnostic tool to link observational data to model data. We will show how future campaigns can make use of the new techniques.

Speaker: Marc Rautenhaus (Universität Hamburg)
• 09:30
Diagnosing Forecast Sensitivity for Field Campaigns using Adjoints 30m

Adjoint models can provide valuable insight into the practical limitations of our ability to predict weather systems, such as extratropical and tropical cyclones, and their associated high-impact weather. An adjoint model can be used for the efficient and rigorous computation of numerical weather prediction forecast sensitivity to changes in the initial state or an earlier point in the forecast. The sensitivity calculations can help to unravel complex instabilities and influences on extratropical and tropical cyclone evolution that occur over a wide range of scales. Adjoints also can be used to explore the rapid growth of small perturbations that lead to large errors on multiple scales and limit the forecast accuracy of high-impact weather events.

In this presentation, we provide highlights from a number of field programs to illustrate the utility of adjoints to: i) inform field program observing strategies; ii) highlight key mesoscale moist instabilities and processes; and iii) diagnose initial condition sensitivities and processes that contribute to forecast errors. We will discuss examples from several fields programs in which we apply NRL’s Coupled Ocean-Atmosphere Mesoscale Prediction System (COAMPS) moist adjoint to diagnose forecast sensitivity including: i) tropical cyclone programs: NASA Hurricane and Severe Storm Sentinel (HS3), NOAA/NASA Sensing Hazards with Operational Unmanned Technology (SHOUT), Office of Naval Research (ONR) Tropical Cyclone Intensity (TCI) and ii) extratropical cyclone programs: North Atlantic Waveguide and Downstream Impact Experiment (NAWDEX) and Atmospheric Rivers Reconnaissance (ARRECON).

The adjoint, tangent linear, and nonlinear models for the nonhydrostatic COAMPS are applied with high-horizontal resolution (5 to 45 km). We show that the initial state sensitivity and forecast errors are well correlated based on results from several different field programs. We compare and contrast results from tropical and extratropical campaigns using several different response functions to explore various aspects of sensitivity and predictability including: i) kinetic energy, ii) accumulated rainfall, iii) integrated vapor transport, and iv) potential vorticity. The adjoint sensitivity results for the extratropical and tropical cyclones underscore the importance of the low- and mid-level moisture distribution and multi-scale interactions. We demonstrate that small perturbations of moisture, winds, and temperature in sensitive regions rapidly evolve into disturbances that impact the predictability of downstream high-impact weather. The forecast sensitivity to diabatic heating is also explored using the adjoint to provide insight into the implications of model error associated with microphysical parameterizations.

The results underscore the need for accurate moisture observations and data assimilation systems that can adequately assimilate these observations in order to reduce the forecast uncertainties for these high-impact events. However, given the nature of the sensitivities and the potential for rapid error growth, the intrinsic predictability for high-impact weather appear to be limited.

Speaker: James Doyle (Naval Research Laboratory)
• 10:00
Impact assessment approaches for field campaign data 30m
Speaker: Mohamed Dahoui (ECMWF)
• 10:30 11:00
Coffee break 30m
• 11:00 13:00
Clouds - Chair: Irina Sandu
• 11:00
Learning about clouds and circulation during EUREC4A 30m

To follow

Speaker: Sandrine Bony (CNRS, LMD/IPSL, Sorbonne University)
• 11:30
Cloudy boundary layers over the Northeast Pacific and Southern Ocean: Field observations and ERA5 30m

In recent years, field campaigns have deployed modern in-situ and remote-sensing instrumentation in diverse marine cloudy boundary layer regimes. For example, the CSET (2015) airborne campaign sampled across the NE Pacific stratocumulus-cumulus transition between California to Hawaii. The Southern Ocean Atmospheric Research program (2016-8) comprised four observational campaigns (SOCRATES, CAPRICORN, MICRE and MARCUS) sampling between Tasmania and Antarctica, including airborne, ship and island measurements of cloud microphysics, precipitation, turbulence, thermodynamic profiles, and aerosols.
This talk will summarize observations taken during these campaigns and compare them with ERA5. In general, ERA5 is remarkably consistent with in-situ wind and temperature measurements, locates high relative-humidity layers accurately, boundary-layer vertical structure moderately accurately, and the vertical structure of cloud, precipitation and ozone somewhat less accurately. It slightly outperforms MERRA-2 in almost all respects.

Speaker: Chris Bretherton (University of Washington)
• 12:00
Field Experiments for NWP: The LITFASS Experience 20m

Numerical weather prediction (NWP) essentially relies on measurements of atmospheric variables for data assimilation, parameterization development and validation. Observatories and supersites nowadays provide comprehensive data sets from operational measurement programs which might be considered as a long-lasting field experiment taking into account the size and variety of measurements. These data are of special relevance for model development as they cover the full spectrum of weather situations and phenomena occurring at a given site over longer periods.
However, additional challenges to the observational capabilities are associated with the increasingly higher spatial resolution of the models and with the parameterization of increasingly complex small-scale physical processes and interactions. These call for data sets which can only be collected within the frame of field campaigns.
Deutscher Wetterdienst (DWD) at its Meteorological Observatory Lindenberg (MOL) in 1995 started the LITFASS (Lindenberg Inhomogeneous Terrain – Fluxes between Atmosphere and Surface: a longterm Study) project, a program to test and to establish a strategy for the determination of soil-vegetation-atmosphere interaction processes over a heterogeneous land surface at the scale of a grid cell of a NWP model. Three major field experiments have been organized within LITFASS around Lindenberg: LITFASS-98, LITFASS-2003, and LITFASS-2009. The basic focus of these experiments was the determination and description of momentum, sensible heat and water vapour fluxes as an area-average at the meso- scale.
The presentation will give an overview on the goals and major results of the three LITFASS field experiments with a special focus on the lessons learned from these campaigns. Another aspect to be mentioned does concern the role of field experiments to test new instruments and measurement strategies which might become operational in future years thus improving the data base for NWP. Ground-based remote sensing systems or unmanned aerial vehicles are prominent examples for such a development.

Speaker: Frank Beyrich (Deutscher Wetterdienst)
• 12:20
FESSTVaL: Field Experiment on sub-mesoscale spatio-temporal variability in Lindenberg 20m

Numerical weather prediction (NWP) models applied on regional scales use a typical grid spacing of O(1 km). While such a grid spacing allows to start explicitly resolving convection - at least deep convection - several features of the flow remain of subgrid-scale nature, e.g. turbulence, shallow convection, or may be distorted by the coarse grid spacing. Large-eddy simulations (LES) with grid spacing of at least O(100 m) can be used to get more information on smaller-scale, generally under-resolved, phenomena. But such simulations also rely on parameterizations, most notably turbulence and microphysics. Getting information on the atmospheric flow on scales O(500 m) from observations remains challenging as the measurement network lacks the spatial resolution. For instance automatic measurement stations of the German Weather Service (DWD) have a typical horizontal distance of O(25 km). This makes the validation of NWP models and LES difficult.

We present the plan for the field campaign FESSTVaL, which deploys a high-density measurement network that will allow us to observe features of the atmospheric flow occurring on scales between 500 m and 5 km. The measurements will be used to (i) improve our process understanding, (ii) validate aspects of convection permitting NWP simulations and (iii) compare different measurement strategies and instrument types, including a citizen science approach and newly available satellite observations, in view of designing appropriate measurement networks of the future. Finally, in addition to the measurements, various simulations will be performed in support of the field campaign and for validation purposes.

With respect to the source of submesoscale variability, the measurement campaign focuses on three different topics: boundary layer patterns, cold pools, and wind gusts. The four topics are inter-connected via cold pools, which both generate boundary layer patterns and wind gusts. Furthermore, usability of citizen-science-based measurements will be investigated. Finally, FESSTVaL will provide a first opportunity to evaluate quality and representativeness of ESA Aeolus products.

The measurement campaign will take place in Lindenberg (east Germany) for an extended summer season in 2020 in the context of HErZ (Hans Ertel Centre for Weather Research). Lindenberg is chosen given the already existing instruments, the support by DWD available on site as well as the relatively flat topography. Moreover Lindenberg experiences more frequent convective activity than many other flat regions in Germany. One particular feature of the planned field experiment is the use of about 100 ground base stations, spread over a domain with a radius of 20km around the Lindenberg observtory.

Speaker: Daniel Klocke (Hans Ertel Centre for Weather Research, DWD)
• 12:40
RALI: the French radar-lidar airborne platform for cloud dynamics and microphysics studies 20m

Mobile Radar-Lidar facilities are unique tools for cloud process analyses and case studies. The radar-lidar airborne platform (RALI) can be deployed on board the French SAFIRE aircraft (Falcon 20 or the ATR42 depending on the targeted areas). RALI consists of a combination of the multi beam 95 GHz Doppler radar RASTA (RAdar SysTem Airborne) and the Doppler high spectral resolution (D-HRS) lidar LNG (Leandre New Generation). Both instruments were developed at LATMOS and DT-INSU (http://rali.projet.latmos.ipsl.fr). LNG operates at three wavelengths (355 nm, 532 nm, 1064 nm), including depolarization and D-HRS at 355 nm. This synergistic platform has been deployed in many field campaigns on SAFIRE aircraft since 2006 combined with radiometry for SW and LW flux measurements (for example AMMA, HYMEX, CHARMEX, HAIC, NAWDEX and EXAEDRE). The unique configuration of the RASTA radar allows for the retrieval of the three-dimensional cloud/precipitation wind above and below the aircraft (data collected during HYMEX data have been recently assimilated in AROME). Ice clouds, water cloud top and aerosol properties (local scale dynamics and radiative parameters) can also be retrieved thanks to the D-HRS lidar. Combination of both instruments give access to an unprecedented set of parameters over the whole atmospheric column. We will present the platform and its capability in terms of microphysical and dynamical processes studies and some applications.

Speaker: Julien Delanoë (IPSL, UVSQ Université Paris-Saclay, Sorbonne Université, CNRS)
• 13:00 14:00
Lunch break 1h
• 14:00 15:30
Working groups Council Chamber, Large Committee Room, Meeting Room 1 ()

### Council Chamber, Large Committee Room, Meeting Room 1

• 15:30 16:00
Coffee break 30m
• 16:00 18:00
Tropical cyclones - Chair: Linus Magnusson
• 16:00
Examples of targeted high-altitude airborne dropsonde deployment strategies for improved tropical cyclone and winter storm prediction 30m

Two recent Tropical Cyclone (TC) Forecast Demonstration Projects (TCFDP) have utilized new and innovative technologies and targeted observing strategies for improving TC track and intensity forecasting: 1) Sensing Hazards with Operational Unmanned Technologies (SHOUT, 2015-16) and 2) East Pacific Origins and Characteristics of Hurricanes (EPOCH, 2017). Both of these projects share the objective of complementing legacy G-IVSP RD-94 dropsonde observational capability of the mid- and upper troposphere with improved NRD-94 mini-dropsonde observing strategy deployed from Global Hawk UAV vehicles in the lower stratosphere.

During the 2018 hurricane season and planned for the 2019 hurricane season is the use of improved versions of the ensemble targeting strategies developed at U. Albany in concert with the U.S. National Hurricane Center and Environmental Modeling center using ECMWF global model together with new instrument design in high-altitude dropsonde deployments from the NOAA GIVSP aircraft. The new targeting strategy and a third generation dropsonde design were implement mid-way through the 2018 hurricane season. Adding to these developments is a second ensemble targeting system utilizing the GFS global model that was developed for use in hurricanes and which was first implemented in the recently completed 2019 Atmospheric Rivers Reconnaissance program for winter storm event forecasting improvement along the U.S. West Coast. This program was directed by the Scripps Center for Western Weather and Water Extremes (C3WE) and conducted jointly with the Air Force Reserve Command (AFRC), NOAA/ Environmental Modeling Center (EMC), National Center for Atmospheric Research (NCAR) and the ONR/Naval Research Lab (NRL). The AFRC/ 53rd Weather Reconnaissance Squadron flew two WC-130J aircraft at maximum altitude, deploying new third generation NCAR/EOL designed, Vaisala produced RD-41 dropsondes.

A new dropsonde targeting strategy was developed at U. Albany using ECMWF ensemble forecasts in the 48-72 hour period to estimate regions of high observational uncertainty for prediction of track and intensity. Midway through the ARR-2019 program a similar product was introduced by EMC using GEFS ensemble global model, allowing consensus uncertainty regions to be identified. The Global Hawk patterns flown in both of these projects used the U. Albany strategy in developing dropsonde deployment locations for each flight. This paper describes how the resulting dropsonde locations compared with these regions of maximum uncertainty. In addition, the locations of these high uncertainty regions relative to key environmental and storm relative features is described and depicted using GOES visible and IR imagery, microwave imagery and concurrent airborne and land-based radar imagery.

Speaker: Peter Black (I.M. Systems Group)
• 16:30
Upper Ocean Data Collection during Operational Hurricane Reconnaissance Missions 30m

In 2011 the U.S. Working Group for Hurricane and Winter Storms Operations and Research approved a multi-year AXBT Demonstration Project to assess whether the collection of upper-ocean temperature observations during operational tropical cyclone (TC) reconnaissance missions could improve coupled numerical model forecasts of TC track and intensity. In 2017, the program was expanded to include salinity observations and was incorporated as a Navy-Air Force partnership in the U.S. National Hurricane Operations Plan. Over the past eight years, with support from the Office of Naval Research, we have collaborated with the U.S. Air Force 53rd Weather Reconnaissance Squadron in the deployment of more than 1000 AXBTs and 80 ALAMO floats in 29 named storms, and with the Naval Oceanographic Office, National Data Buoy Center, and Naval Research Laboratory – Monterey in the transmission, assimilation, and analysis of the observations and impact. We have also begun research collaborations with scientists from ECMWF and the UKMET Office.

Achievements and results to date include near-real-time assimilation of upper-ocean temperature and salinity observations, increased accuracy of ocean model forecasts, improved track and intensity forecasts in coupled dynamical models, and development of a targeting system to identify critical ocean observation areas ahead of tasked tropical cyclone missions. Challenges include understanding when and where the observations are most impactful, and identifying an optimal sampling strategy to facilitate transition to an operational program. Topics in focus here will include the operational construct of the research program, an overview of the observations collected to date, a discussion of current objectives and future plans, and an invitation to collaborate with workshop participants to maximize the effectiveness of future observations.

Speaker: Elizabeth Sanabia (U.S. Naval Academy)
• 17:00
New observational strategies for operational global and hurricane model improvements using airborne Tropical Cyclone and Winter Storm surveillance flights 30m

A new aircraft observing strategy is proposed for obtaining much-needed atmospheric dropsonde observations throughout the entire depth of the troposphere within the inner core of Tropical Cyclones (TCs) and their environment during TC season, as well as over developing winter storm systems in the Central and Eastern Pacific (CPAC and EPAC) upstream from the U.S. West Coast Atmospheric River (AR) landfalls, Central Plains severe weather and rapidly developing East Coast winter storms. The observing strategy involves new and improved dropsonde instruments deployed utilizing new and improved dropsonde targeting strategies which define regions of model uncertainty using operational global ensemble forecast models. Data from these flights would be for the purpose of reducing model uncertainty in the prediction of TC track and intensity as well as location and intensity of winter storm systems. Secondarily, initial analyses suggest that data from these flights would play a role in improving longer-range, down-stream forecasting of TC’s such as Extratropical Transition (ET) and storms in other global basins as well as severe weather impacts over the CONUS region during the winter season. The use of emerging new dropsonde targeting techniques defining regions of operational model uncertainty is critical to the successful implementation of this new plan. Initial experimentation suggest that use of multiple global models (and their ensembles) with different uncertainty estimation methods helps to further reduce targeting uncertainty and maximize the use of dropsonde observations in the operational data assimilation system.

Preliminary results from targeting strategies implemented with the recent hurricane seasons using Global Hawk and G-IV aircraft will be discussed with respect to impacts on the new NCEP operational Global Forecast System (GFS) based on the Finite Volume Cubed Sphere (FV3) dynamic core, and operational Hurricane Weather Research and Forecast (HWRF) models. These strategies were further exploited during the Atmospheric Rivers Reconnaissance campaign in February 2019 using dual WC-130J aircraft. NWS has partnered with CW3E, Navy, NCAR and SUNY Albany in providing ensemble model based targeting strategies for dropsonde deployments for the ARR-2019 as a pilot project. Preliminary results from this campaign will be presented. Impact on numerical model guidance, local forecasts and downstream impacts will await further in-depth study.

Speaker: Vijay Tallapragada (NOAA/NWS/NCEP/EMC)
• 17:30
Analysis and forecast using dropsonde data from inner-core region of tropical cyclones obtained during the aircraft missions of T-PARCII 30m

The inner core of tropical cyclone (TC) Lan was observed on 21-22 October 2017 by newly developed GPS dropsondes during the aircraft missions of the Tropical Cyclones-Pacific Asian Research Campaign for the Improvement of Intensity Estimations/Forecasts (T-PARCII). On 25-28 September 2018, the inner core of TC Trami was also observed by T-PARCII team with the support of Science and Technology Research Partnership for Sustainable Development (SATREPS). So far, the eyewalls were penetrated nine times with a Gulfstream II jet and 90 dropsondes were dropped from 43,000 ft. The estimated minimum sea-level pressure was 925 hPa in the aircraft missions for TC Lan, while it was 920 hPa for TC Trami. From 2018, we started to transmit the BUFR data on GTS through Japan Meteorological Agency (JMA). To evaluate the impact of dropsondes on forecast skill, the forecast experiments were conducted using a JMA non-hydrostatic model (JMA-NHM) with a JMANHM-based mesoscale four-dimensional data assimilation system with a grid-spacing of 5 km for TC Lan. Then, we evaluated the forecast skill against the best track data published by the Regional Specialized Meteorological Center (RSMC) Tokyo. Track and heavy rainfall forecast skills generally improved by about 10 % with the assimilation of the dropsonde data, while the intensity forecasts were generally degraded. The degeneration of the intensity forecast skill is, however, potentially due to uncertainties in the best track data as the best track data set usually relies on the Dvorak technique involving the error of the order of 10 hPa. The benefits of inner-core observations described are encouraging, yet at the same time they remind us of the importance of the ground truth in the researches of TC forecasting. Other relevant researches including the assimilation with a JMA global data assimilation system and sensitivity analysis will be also presented.

Speaker: Kosuke Ito (University of the Ryukyus)
• Wednesday, 12 June
• 09:00 10:30
Tropical and extra-tropical dynamics - Chair: Andreas Schäfler
• 09:00
The impact of dropsonde and extra radiosonde observations from the field campaigns NAWDEX and SHOUT in 2016 30m

Dropsonde observations from three research aircrafts in the north-Atlantic region as well as several hundred additionally launched radiosondes over Canada and Europe were collected during the transatlantic field campaign NAWDEX in autumn 2016. In addition, over 500 dropsondes were deployed during NOAA’s SHOUT and Reconnaissance missions in the west-Atlantic basin, complementing the conventional observing network for a total of 13 intensive observation periods. This unique dataset was assimilated within the framework of cycled data denial experiments performed with the global model of ECMWF.
On average, these additional observations led to a reduction in forecast error of a few percent in a large area covering the North Atlantic and Europe. The error reduction mainly seems to be related to three particular sensitive episodes that are associated to the extratropical transitions of tropical storm Karl and hurricanes Matthew and Nicole. The forecast sensitivity to observations impact (FSOI) also exhibits largest beneficial impacts for dropsondes near tropical cyclones, followed by dropsondes over the North Atlantic and additional Canadian radiosondes.

Speaker: Martin Weissmann (DWD/LMU)
• 09:30
Atmospheric River Reconnaissance to Improve Forecasts: Needs, Approach and Underlying Science 30m

The Atmospheric River Reconnaissance project “AR Recon” formulated a targeting method focused on AR landfall prediction on the U.S. West Coast, where AR landfall position forecast errors at 1-4 days lead time range from 200-400 km on average (Wick et al. 2013, DeFlorio et al. 2018), and can contribute to significant errors in extreme precipitation forecasts (e.g., Ralph et al. 2010, 2011). The recent addition of moist processes in an adjoint method concluded that errors in the location and characteristics of ARs offshore as the leading source of initial condition error for landfalling storm forecasts on the west coast (Doyle et al. 2014; Lavers et al. 2018; Reynolds et al. 2019). These forecast errors impact water decisions in the West, including those associated with mitigating flood risk and drought (http://cw3e-web.ucsd.edu/firo/).
The AR Recon project is a multi-year, interagency, cooperative effort to collect unique dropsonde observations in and around ARs off the U.S. West Coast to improve AR-landfall-associated weather forecasts during the cool season. It has collected data with multiple aircraft in 3 ARs in February 2016 (two Air Force C-130s), 6 ARs in January-February 2018 (involving a mix of two Air Force C-130s, and NOAA’s G-IV), and 6 ARs in February 2019 (used two Air Force C-130s). In 2019, AR Recon also supported the deployment of additional drifting buopys, with surface pressure sensors, in the northeast Pacific. Additionally, airborne GPS met observations have been made in some cases (Haase).
Global modeling centers (NCEP, US Navy, ECMWF), and regional modelling efforts (COAMPS; West-WRF) have teamed up to collaborate. An AR Data Assimilation and Modeling Steering Committee (the co-authors of this abstract) has brought together diverse expertise and substantial institutional capacity to carry out the collaboration.
This presentation will present a status report on data collection and analysis.

Speaker: Vijay Tallapragada (NOAA/NCEP)
• 10:00
The gauging and modelling of rivers in the sky 30m

-

Speaker: David Lavers (ECMWF)
• 10:30 11:00
Coffee break 30m
• 11:00 12:10
Tropical and extra-tropical dynamics - Chair: David Lavers
• 11:00
Forecast products for flight planning from a researchers' perspective 30m

The scientific aim of the North Atlantic Waveguide and Downstream Impact Experiment (NAWDEX; Schaefler et al. 2018) was to increase the physical understanding and to quantify the effects of diabatic processes on disturbances of the jet stream and their influence on downstream high-impact weather. The field campaign in September/October 2016 involved four research aircraft which performed in total 49 research flights. In preparation of NAWDEX and previous related field campaigns such as T-NAWDEX Falcon (Schaefler et al. 2014) and ML-Cirrus (Voigt et al. 2017), researchers at ETH Zurich developed web-based forecasting tools using output of ECMWF’s high-resolution and ensemble forecasting system. These tools allowed to identify regions where diabatic processes were at work and where they interacted with the jet stream. For example, the computation of warm conveyor belt trajectories from the ensemble forecasting system helped to assess the probabilities of these rapidly ascending air streams and to design preliminary flight plans a few days in advance. The flight plans were then refined using the high resolution deterministic forecast. The use of web-based vertical cross-section tools as well as the computation of trajectories starting from the envisaged flight path were particularly helpful. Accordingly, this talk will demonstrate the variety of forecasting products and will provide examples on how they were used during NAWDEX and related campaigns.

Speaker: Julian Quinting (Karlsruhe Institute of Technology)
• 11:30
Comparison between airborne remote sensing observations of an extratropical cyclone and short-term forecasts using a hierarchy of models. 20m

The main goal of the study is to analyze the representation of the Stalactite Cyclone (29 September-2 October 2016) in a hierarchy of models (mesoscale NWP model, global NWP model, climate models). The Stalactite Cyclone is an extratropical cyclone that has been intensively observed during the international field campaign NAWDEX. Its later stage of development is associated with the onset of a low-predictable European block event. Short-term forecasts of all the models are compared with observations made with the RALI platform on board the SAFIRE Falcon. The RALI platform is composed of a Doppler radar RASTA and a three-wavelength backscatter lidar LNG. Two flights of the SAFIRE Falcon have been performed in the ascending and outflow regions of the warm conveyor belt of the cyclone. The comparison between the observations and the model outputs mainly relies on the horizontal wind speed derived from the Doppler radar RASTA, the radar reflectivity and the ice water content retrieved from the RALI observations. The aim of all these observations is to provide indirect information about the diabatic heating rates along the warm conveyor belt and their impact on the dynamics through modification of potential vorticity. The ability of the models to accurately represent the diabatic heating rates, the potential vorticity and the wind within the Stalactite Cyclone and its downstream influence on the blocking formation is assessed.

Speaker: Gwendal Riviere (LMD/IPSL, ENS, PSL Research University, École Polytechnique, Université Paris Saclay, Sorbonne Universités, CNRS)
• 11:50
ASKOS-WIND – A Contribution to the Aeolus CAL/VAL Campaign in Cape Verde in June-July 2020 20m

In August 2018, the Aeolus satellite carrying the first UV Doppler lidar in space (ALADIN) was successfully launched. The particular gap that Aeolus is closing in the global observing system is measurements of winds in cloud free regions and thus we expect Aeolus to substantially improve analysis fields and subsequently predictions of synoptic- to planetary-scale wave phenomena in the Tropics. As part of the Aeolus CAL/VAL activities, an experimental campaign named ASKOS will be organized in June-July 2020 in Cape Verde. ASKOS will deploy advanced instrumentation over Cape Verde to provide unprecedented observations of high quality and accuracy for the wind and aerosol component of Aeolus.

Cape Verde during boreal summer is ideal for this study. The generally high aerosol loading is interesting because it will allow the measurement of both aerosol optical properties and wind, thus opening the way to the study of the interaction between the two. The midlevel African easterly jet allows for the formation of synoptic-scale African easterly waves (AEWs) that typically reach their maximum intensity close to the coast of West Africa. AEWs interact with convection and its mesoscale organisation through modifications in wind, temperature and vertical wind shear, and often serve as initial disturbances for tropical cyclogenesis. In addition, the tropical atmosphere sustains different types of planetary waves that frequently interact with the monsoon and AEWs.

Science questions to be addressed in ASKOS-WIND include: (A) How well does Aeolus monitor winds at different vertical levels in comparison with aircraft measurements and what limits the quality of the retrievals? (B) How well are characteristics of wave disturbances represented in analysis and forecast data relative to the satellite measurements ? (C) How much deterioration do we get if we deny the satellite / aircraft measurements to the data assimilation system? (D) Does a better analysis lead to better forecasts of waves rainfall, dust emission and tropical cyclogenesis?

Speaker: Peter Knippertz (Karlsruhe Institute of Technology)
• 12:15 13:00
Poster session Conference/G-WR - Weather room

### Conference/G-WR - Weather room

#### ECMWF

60
• 13:00 14:00
Lunch 1h
• 14:00 16:00
Polar processes - Chair: Jonathan Day
• 14:00
Improving NWP models with observations from the Iceland Greenland Seas Project 30m

The Iceland Greenland Seas Project (IGP) is a coordinated atmosphere-ocean research program investigating climate processes in the source region of the densest waters of the Atlantic Meridional Overturning Circulation. During February and March 2018, a field campaign was executed over the Iceland and southern Greenland Seas that utilized a range of observing platforms to investigate critical processes in the region – including a research vessel, a research aircraft, moorings, sea gliders, floats and a meteorological buoy. A remarkable feature of the field campaign was the highly-coordinated deployment of the observing platforms; the research vessel and aircraft tracks were planned in concert to allow simultaneous sampling of the atmosphere, the ocean and their interactions. Here we use some of these observations to evaluate the quality of a number of numerical weather prediction forecasts, analyses and reanalyses. In particular we make use of low-level aircraft observations where bulk and turbulent atmospheric variables are available. A coupled evaluation, using boundary-layer observations from both the atmosphere and ocean is underway.

Speaker: Ian Renfrew (University of East Anglia)
• 14:30
What will happen during YOPP SOP3? 30m

The Special Observing Periods (SOPs) within the Year of Polar Prediction aim to provide enhanced observations for the benefit of model improvement for NWP. SOP1 and SOP2 provide additional radiosoundings for a winter and summer period during 2018. For the third one, the SOP3, it is the ambition to coordinate additional observations complementary to the MOSAiC effort. The aim is to target warm airmasses that are heading to the Arctic and the MOSAiC site and cold airmasses advected southward. For the SOP3 to be successful, NWP support is needed for planning. The presentation will present the planning and how NWP centers can be of support and what the additional data can provide to the model development.

Speaker: Gunilla Svensson (Stockholm University)
• 15:00
Arctic Clouds - Evaluating Modelled Cloud with Field Observations 30m

Extensive surface based remote sensing observations of Arctic clouds have been made during two recent research cruises: Arctic Cloud in Summer Experiment (ACSE, 2014) and the Microbiology-Ocean-Cloud Coupling in the High Arctic (MOCCHA, 2018). The cloud properties were retrieved using Cloudnet from measurements by Doppler cloud radar, lidar, scanning microwave radiometer, and 6-hourly radiosondes. Here we present details of the observations and a preliminary evaluation of model output from the IFS.

Speaker: Ian Brooks (University of Leeds)
• 15:30
YOPP supports the Japanese Arctic research cruise in 2018 30m

Unusual states of the Arctic regions, for example, less sea-ice extent, high temperatures in the atmosphere and ocean, much snowfall, and extreme weather events in the Arctic and beyond, have been prominent in recent years in particular during winter time. Those phenomena are scientifically important for understanding the air-ice-sea coupled physical processes and improving skills of numerical models. Skillful forecasts of weather, sea ice and ocean are useful for ship navigation as well as activities of indigenous people. The Year of Polar Prediction (YOPP) proposed by the World Meteorological Organization (WMO) Polar Prediction Project (PPP) provides a great opportunity to collaborate with observing and modeling platforms. In November 2018, the Japanese research vessel Mirai went to the southern Chukchi Sea to observe the unusual conditions of the atmosphere, sea ice, and ocean during the beginning of the freezing. To succeed this unique cruise, skillful forecasts were vital for this ice-strengthen ship. This presentation gives an overview of this cruise from the viewpoint of polar predictions.

Speaker: Jun Inoue (National Institute of Polar Research)
• 16:00 16:30
Coffee break 30m
• 16:30 18:00
Working groups Council Chamber, Large Committee Room, Meeting Room 1 ()

### Council Chamber, Large Committee Room, Meeting Room 1

• 19:30 21:00
Workshop dinner at Bel & The Dragon
• Thursday, 13 June
• 09:00 10:10
Stratosphere, mountains and wind - Chair: James Doyle
• 09:00
TEAMx: A coordinated effort to investigate transport and exchange processes in the atmosphere over mountains 30m

Mountains have a profound impact on synoptic- and meso-scale atmospheric processes. They also shape the transfer of heat, momentum and mass (water or trace gases) between the ground, planetary boundary layer and the free atmosphere. An integral part of past international research programmes that focused on the impact of mountains on the atmosphere (e.g., ALPEX, PYREX and MAP) was a deployment of special observing facilities in large-scale field campaigns. Significant progress in understanding and prediction of processes in and around complex terrain ensued, for instance in relation to gravity-wave-induced phenomena and orographic precipitation. In the two decades since MAP, technological and scientific progress has extended the range of phenomena that can be accurately observed and modelled towards smaller spatial scales. This forms the basis for an internationally coordinated program to study exchange processes over mountains, their interaction with mesoscale processes and their role in the climate system.

From the experimental perspective, terrain heterogeneity creates challenges in practical use and interpretation of observations. Among others, these include limited representativeness of point measurements, special requirements for data post-processing (e.g., for turbulence measurements), and limited validation of satellite remote-sensing retrieval algorithms over regions of complex terrain. From the modelling perspective, sub-grid-scale orographically-induced processes are typically not accounted for in parameterizations of land-surface exchange, planetary-boundary-layer turbulence and convection. Both numerical weather prediction and climate change simulations suffer from model errors caused by imperfect representation of the flow over mountains. These errors remain one of the major sources of uncertainty for the Earth System models despite the ever-increasing model resolution.

This contribution offers an overview of TEAMx (Multi-scale Transport and Exchange Processes in the Atmosphere over Mountains – programme and experiment), a recently initiated programme focusing on the investigation, field observations, and numerical modelling of exchange processes between mountainous terrain and the free atmosphere.

Speaker: Vanda Grubišić (National Center for Atmospheric Research)
• 09:30
Global forecasts of atmospheric gravity waves for observational campaigns 20m

The Institute of Atmospheric Physics of the German Aerospace Center (DLR) organized and participated in three observational campaigns on atmospheric gravity waves (GWs) in the past 7 years: GW-LCYCLE I (Kiruna, Sweden, 2013), DEEPWAVE (Christchurch, New Zealand, 2014) and GW-LCYCLE II (Kiruna, Sweden, 2016). The overreaching goal of all the campaigns with combined airborne and ground-based measurements was a better understanding of the GW sources in the troposphere and lower stratosphere and the wave propagation to the middle and upper atmosphere. The Scandinavian mountain range and the Southern Alps of New Zealand are hotspots of stratospheric mountain wave activity. However, the region of the Andes and the Antarctic Peninsula is the global hotspot in terms of stratospheric GW activity and momentum fluxes. Hence, the investigation of GWs in this region is part of the upcoming SOUTHTRAC campaign in September 2019.

High-resolution operational IFS forecasts (HRES) of the ECMWF have been used for mission and flight planning. Hereby, standardised products on a mission web site as well as an optimized representation in the mission support system (Rautenhaus et al., 2012) were employed. Additionally, HRES forecasts serve as valuable information for controlling autonomous ground-based measurements.

Because of their high fidelity, operational IFS analyses have been often used for post-campaign investigations, e.g., as an overview of the meteorological background conditions, see Gisinger et al., 2017. IFS data were used for the direct comparison or combination of model and observational data (e.g., Ehard et al., 2018) and for interpretation of the results. We found a remarkable agreement of the simulated wave structure in the IFS short-term forecast and space-borne lidar observations (Dörnbrack et al. , 2017). This indicates that the finer resolution and increasing realism of operational NWP model outputs offers a valuable quantitative source for mesoscale flow components which were hitherto not accessible globally.

At the workshop, we want to show briefly how we make use of ECMWF forecasts during a gravity wave campaign. We want to present some selected results of the past campaigns. This will include not only case studies of individual GW events but also long-term (1 year) comparison between model and lidar temperature data
of the middle atmosphere at Rio Grande (Argentina).

Dörnbrack, A., S. Gisinger, M. C. Pitts, L. R. Poole, and M. Maturilli, 2017: Multilevel Cloud Structures over Svalbard. Mon. Weath. Rev., 145 (4), 1149-1159, doi:10.1175/MWR-D-16-0214.1

Ehard, B., S. Malardel, A. Dörnbrack, B. Kaifler, N. Kaifler, and N. Wedi, 2018: Comparing ECMWF high resolution analyses to lidar temperature measurements in the middle atmosphere. Q. J. R. Met. Soc. 144, 633-640. doi:10.1002/qj.3206

Gisinger, S., A. Dörnbrack, V. Matthias, J. D. Doyle, S. D. Eckermann, B. Ehard, L. Hoffmann, B. Kaifler, C. G. Kruse, and M. Rapp, 2017: Atmospheric Conditions during the Deep Propagating Gravity Wave Experiment (DEEPWAVE), Mon. Wea. Rev., 145, 4249-4275, doi:10.1175/MWR-D-16-0435.1

Rautenhaus, M., Bauer, G., and Dörnbrack, A., 2012: A web service based tool to plan atmospheric research flights, Geosci. Model Dev., 5, 55-71, doi:10.5194/gmd-5-55-2012

Speaker: Isabell Krisch (DLR)
• 09:50
The Strateole-2 long-duration balloon project in the deep tropics: benefiting from and improving weather forecasts? 20m

Three long-duration stratospheric balloons were released in February 2010 from Seychelles Island (5°S) by the French space agency (CNES), within the pre-Concordiasi campaign. Once at their float altitude at $\sim$20 km, these balloons drift on constant-density surfaces, and are simply advected by the wind. The pre-Concordiasi flights lasted for 3 months each. In-situ meteorological measurements (temperature, pressure, and wind deduced from successive balloon positions) were performed every 30 s during the flights, and have revealed large discrepancies between observed winds and those in analyses issued by various operational centers for time periods as long as 1~month (Podglajen et al., 2014). The errors in modeled winds have been primarily associated with Kelvin and Rossby-gravity wave packets that were not captured in the analyses, despite their planetary-scale structure. The largest errors occurred over the Indian and Eastern Pacific oceans, where in-situ wind observations are very scarce.

These results contributed to the motivation of the forthcoming Strateole-2 balloon campaigns, which will release $\sim$ 50 similar long-duration balloons in the deep tropics in the 2019-2024 time frame. Strateole-2 balloons will circum-navigate
the Earth around the equator in the lower stratosphere, and provide observations over both continents and oceans. Meteorogical measurements performed during the flights will be sent on the GTS so as to be assimilated by numerical weather
prediction (NWP) systems. The wind observations should particularly contribute to improve NWP wind analyses and forecasts in the tropics. On the other hand, better operational wind products are also very useful for the campaign management. They
for instance provide (i) better guidance on future balloon trajectories, which are monitored for safety reasons, and (ii) opportunities to manage coordinated measurements with instrumented sites at ground.

The presentation will briefly review the past pre-Concordiasi experience, and provide a detailed view of Strateole-2 and its potential interactions with the operational weather forecast community.

Speaker: Albert Hertzog (LMD/IPSL)
• 10:10 10:40
Coffee break 30m
• 10:40 11:20
New observational capabilities - Chair: Emma Pidduck
• 10:40
The CNES stratospheric balloon activities: capabilities, mission and operations 20m

For more than 50 years of experience, the Centre National d'Etudes Spatiales (CNES) has been supporting scientific ballooning, which remains the most cost effective means to access to near space science.
This paper will give a quick overview of the CNES capabilities and services for operational balloon activities: Zero Pressure Balloon, Super Pressure Balloon and Sounding balloons. It will focus on the on-going development to improve them. The most recent balloon campaigns will be presented, as well as the future campaigns.

Speaker: André Vargas (Centre National d'Etudes Spatiales)
• 11:00
Saildrone: A global class Unmanned Surface Sailing Vehicle for air-sea interaction observation and its potential as a reliable data source for NWP models 20m

Over the past decade, significant progress has been made in the global ocean observing system (GOOS), which is monitoring much of the upper ocean on a global scale in real time with multiple observing platforms. But observation of air-sea fluxes has been relying on fixed surface moored buoys and research and voluntary ships with limited spatial coverage. As a result, the current GOOS is not able to adequately observe the air-sea interaction processes across fronts and eddies. A recent technology development, the Saildrone, is an Unmanned Surface Vehicle (USV) powered by wind and solar energy with a range of more than 6,000 nautical miles, making it a potential platform to sample across fronts and weather systems over the global ocean. To make the Saildrones capable of observing air-sea interaction processes, we have installed sensors with equivalent or better quality than those currently used on Tropical Atmosphere and Ocean (TAO) buoys for air-sea flux measurements, and a 300-kHz Acoustic Doppler Current Profiler for upper ocean current measurements. So far, two pilot Saildrone missions have been completed in the tropical Pacific, as part of the Tropical Pacific Observing System (TPOS)-2020 project: one mission with two Saildrones deployed and recovered at San Francisco, California; the other one with four Saildrones deployed and recovered at Honolulu, Hawaii. Both missions reached the equator and sampled across the tropical Pacific cold tongue fronts and oceanic vortices. We will present the Saildrone data from these TPOS missions and use these results as an example to demonstrate the potential of Saildrones for air-sea interaction studies over the global ocean. While the Saildrone observations can provide the needed data to the Numerical Weather Prediction (NWP) models, the NWP forecasts are essential for the planning and execution of Saildrone missions for navigation and better sampling strategies. Through this presentation, we hope to open the dialogue between the USV observation and NWP modelling communities to facilitate future collaborations in better collection and use of the air-sea interaction data collected by these new platforms.

Speaker: Dongxiao Zhang (JISAO/University of Washington and NOAA/Pacific Marine Environmental Laboratory)
• 11:30 13:00
User voice corner - Can ECMWF better support field campaigns? Council Chamber, Lecture Theatre, Meeting Room 1 ()

### Council Chamber, Lecture Theatre, Meeting Room 1

• 13:00 14:00
Lunch break 1h
• 14:00 16:00
Working groups summary and discussion Lecture Theatre ()