BoBBLE: Bay of Bengal Boundary Layer Experiment

Lead Research Organisation: University of Reading
Department Name: Meteorology


The South Asian summer monsoon (June-September) provides 80% of the annual rainfall for over one billion people, many of whom depend on monsoon rains for subsistence agriculture and freshwater. It is critical to forecast accurately not only the seasonal rainfall, but also rainfall variations within the summer. Sub-seasonal "active" and "break" phases can last weeks, resulting in floods and droughts across broad areas of South Asia.

Air-sea interactions are key to understanding and predicting monsoon behaviour. Ocean surface temperatures in the Bay of Bengal, east of India, remain very warm (above 28 C) throughout the summer. Evaporation from the Bay provides moisture and energy to monsoon depressions that form over the Bay and bring substantial rain to India. It is not understood how the Bay remains warm despite losing energy to these systems. Ocean temperature and salinity variations across the Bay are known to drive changes in rainfall over the Bay and surrounding land, but it is not clear how these arise or how they are maintained. This is particularly true for east-west variations in the southern Bay, a focus of this project. Although air-sea interactions are important to the monsoon, weather predictions are made with models of only the atmosphere. There is potential to improve monsoon forecasts by including well-represented air-sea interactions in models.

The Bay of Bengal Boundary Layer Experiment (BoBBLE) proposes an observational campaign for the southern Bay, during the established monsoon (mid-June to mid-July). BoBBLE will deploy two ships, six ocean gliders and eight floats to collect an unprecedented range of oceanic and air-sea flux observations. The ships will occupy locations in the southwest and southeast Bay, as well as tracing east-west and north-south paths between those locations, measuring ocean temperature, salinity and currents. Two gliders (automated underwater vehicles) will accompany each ship, with two others between the ships, diving to 500 metres every 2 hours to measure temperature, salinity and currents. Diurnal variations in air-sea fluxes and ocean temperatures may affect the development of weather systems. A novel configuration of the gliders will allow computations of horizontal transports of heat and salt. The floats (automated submersibles) will be deployed in the Bay to measure the ocean to 2000 metres every 5 days. They will remain in the Bay after BoBBLE, enhancing the observing network. Ships and gliders will also measure ocean chlorophyll, which absorb sunlight and alter near-surface ocean temperature, influencing air-sea interactions.

BoBBLE scientists will analyse these observations, along with routine datasets, to understand the evolution of conditions in the Bay and how they influence the atmosphere. Particular emphasis will be placed on estimating the uncertainty in existing datasets of air-sea fluxes by validating them against all available observations. The best-performing datasets will be used to improve estimates of air-sea exchanges and their variability on daily to decadal timescales, to calculate budgets of heat and freshwater fluxes in the Indian Ocean and the Bay, and to validate model simulations.

A hierarchy of model simulations will reveal how conditions in the Bay are maintained and how air-sea interactions influence the monsoon. Simulations with an ocean model, forced by and validated against BoBBLE observations, will isolate the roles of air-sea fluxes (including the diurnal cycle), chlorophyll and horizontal transports in maintaining and recharging ocean structure after each weather system passage. Retrospective forecasts of the BoBBLE period with atmosphere-only and atmosphere-ocean coupled models will demonstrate how air-sea interactions influence monsoon rainfall predictions. Multi-decadal simulations will evaluate how air-sea interactions and coupled-model systematic errors influence daily-to-seasonal monsoon variability.

Planned Impact

Here we summarise who is likely to benefit from the BoBBLE research and how they will benefit. The Pathways to Impact attachment describes what will be done during BoBBLE to maximize the likelihood of the BoBBLE research reaching those beneficiaries.

Academic researchers in the fields of climate science, meteorology and oceanography will benefit from the advances made and also the new techniques (Seagliders, Argo floats, numerical modelling schemes) that BoBBLE will develop and apply. Some of these techniques will be specific to the local challenges of monsoon system in the Bay of Bengal, but others will be globally applicable. The focus of BoBBLE is on the physics of air-sea interaction, but both the observations and the modelling will include elements of biogeochemistry, so our results will benefit those working on fishery management, carbon sequestration and ecosystem modelling. Ocean scientists in the Bay of Bengal region will benefit by learning how to deploy, operate and pilot Seagliders and how to best design cost-effective surveys to obtain near-surface multidisciplinary data.

Weather forecasters both in government agencies and commercial companies around the world will benefit from BoBBLE, directly through provision of novel data sets in a rarely-observed region, and indirectly through the improvements in climate and weather forecasting techniques and understanding. Fundamentally, BoBBLE is designed to deliver improvements to the forecasting of the monsoon rainfall, and thus will benefit those in Southeast Asian countries tasked with management of water resources. The Indian NCMRWF will benefit from improved modelling parameterisations, and the UK Met Office will benefit from improved understanding of monsoon triggers contributing to improving the HadGEM and UM family of models. This contributes directly to increasing the value of weather/climate forecasting products that they can sell.

The PDRAs working on the project will gain valuable skills in numerical modelling, seagoing oceanography and meteorology, data processing and analysis, Seaglider operation and piloting, and science communication. They will also gain transferable skills including time management, networking, project management, leadership, communication to different audiences, and computer programming. These will be valuable skills should they move into business or industrial careers.

In the longer term, we will make a contribution to resolving the skill shortage in UK scientists and technologists, by attracting further young people into science and into tertiary education through our outreach and media efforts. Norfolk in particular is a region where take-up of tertiary education is low, so we will focus our efforts to raise ambitions and confidence in East Anglia.
Description The University of Reading component of the Bay of Bengal Boundary Layer Experiment (BoBBLE) has made substantial advances in our understanding and prediction of "active" and "break" periods in Indian summer monsoon rainfall. Active periods are 1-3 week spells of above-normal rainfall over India, which are associated with floods; break periods are 1-3 week spells of below-normal rainfall associated with droughts. These periods are linked to large, organised areas of tropical thunderstorm activity that form in the Indian Ocean and move north and east over India and the Bay. Our work focused on the ability of weather and climate models to simulate these events, particularly whether interactions between the atmosphere and the oceans surrounding India were important for predicting these events.

We found that the Met Office climate model, used for seasonal forecasting and climate-change projections, has is able to represent these active and break events well. Including feedbacks between the atmosphere and ocean is important for simulating these events and their effect on rainfall over India. The horizontal resolution (grid spacing) of the model is also important for capturing these events; we recommend 90 km grid spacing or finer. We also found that 90km resolution is essential for capturing monsoon depressions, which are cyclonic storms that can bring extreme rainfall to India. These depressions are more common in monsoon active periods.

We developed a version of the Met Office weather forecasting model that includes interactions between the atmosphere and the ocean. The standard version the Met Office use for global weather forecasts does not include these interactions. We tested whether our "coupled" configuration produced better forecasts over India than the standard version, using the summer 2016 monsoon season -- the period of the BoBBLE field campaign -- as a case study. We found that our coupled configuration improved the predicted timing of active and break spells over India, particularly for the end of these spells. However, we found no other systematic changes in the forecasts, which suggests that the effects of atmosphere-ocean feedbacks on short-range weather forecasts are subtle, though potentially important for the predicting the transition between active and break periods. Our forecasts were performed at the relatively coarse resolution (17km) used for global weather forecasting, not the finer resolution (4km) typically used for regional weather forecasts or predictions of extreme events (e.g., tropical cyclones). Hence, our results raise important new questions about the effects of atmosphere-ocean coupled interactions on higher-resolution (convective-scale) weather forecasts. We are investigating these questions in a new research project, CASPER, funded by the Newton Fund, in collaboration with our Indian partners.

We have collaborated with the Indian National Center for Medium-Range Weather Forecasts (NCMRWF) on coupled atmosphere-ocean prediction of the Indian monsoon. NCMRWF use the Met Office weather forecasting model operationally. We worked together with NCMRWF to transfer our coupled version of the Met Office model to India, where it is being evaluated more extensively for operational implementation for global and regional weather forecasts.
Exploitation Route Our results on the effects of air-sea interactions on monsoon weather prediction could be taken forward by the Met Office and by NCMRWF, as they consider whether to perform global and regional weather forecasts with coupled atmosphere-ocean models, or continue to use only atmospheric models that do not consider air-sea feedbacks. Our results show that, for global-scale weather forecasts, air-sea feedbacks have a limited role in monsoon weather prediction, though there are clear benefits to predicting the timing of active and break spells.

Our air-sea coupled configuration of the Met Office model could be taken forward by the Met Office and NCMRWF for potential operational implementation in global and regional weather forecasting. We are working with the Met Office and NCMRWF to evaluate this coupled configuration for high-resolution (4km) regional weather forecasts for South Asia, particularly for tropical cyclones and related extreme weather systems (e.g., monsoon depressions).

Our outcomes on the ability of the Met Office model to predict active and break spells could be taken forward by the agricultural or energy sectors, to inform planning decisions for planting and harvesting, or for dam management for hydropower generation and freshwater supplies.
Sectors Agriculture, Food and Drink,Energy,Environment

Description Coupled Air-Sea Prediction of Extreme Rainfall (CASPER)
Amount £837,979 (GBP)
Organisation Newton Fund 
Sector Public
Country United Kingdom
Start 09/2019 
End 03/2021
Description Seasonal Teleconnections to the Indian Monsoon: Understanding the Large-scale Atmospheric Tropical Environment (STIMULATE)
Amount £822,360 (GBP)
Organisation Newton Fund 
Sector Public
Country United Kingdom
Start 09/2019 
End 03/2021
Title Mixed-layer coupled configuration of Met Office Numerical Weather Prediction Model 
Description We have created a configuration of the Met Office global numerical weather prediction (NWP) model that is coupled to a mixed-layer ocean model. The ocean model has been developed by our research team. This air-sea coupled NWP model enables weather forecasts to represent the influence of atmospheric variability on the underlying ocean, and the consequent impacts of changes in ocean surface temperatures on the atmosphere. Most NWP models neglect changes in the ocean, as they were previously believed to be unimportant on the 7-10 day timescales used for NWP. However, our research and that of other scientists has shown that changes in ocean surface conditions may be important for the development of high-impact weather systems, particularly tropical weather systerms such as the monsoon depressions of interest to BoBBLE. To test the influence of air-sea interactions on NWP for the South Asian monsoon, we are re-forecasting the 2016 summer monsoon season with our air-sea coupled NWP configuration. The output will be compared to the operational (atmosphere-only) predictions issued in 2016, to understand whether air-sea coupling affects NWP for high-impact monsoon weather systems. 
Type Of Material Improvements to research infrastructure 
Year Produced 2017 
Provided To Others? No  
Impact Research is ongoing. We will update this record when conclusions from our re-forecasts of the 2016 monsoon season are identified. 
Description IISc Bangalore collaboration for BoBBLE 
Organisation Indian Institute of Science Bangalore
Department Centre for Atmospheric and Oceanic Sciences
Country India 
Sector Academic/University 
PI Contribution Through the BoBBLE project, we are collaborating with the Indian Institute for Science, Bangalore, on analysis of BoBBLE field campaign data and related model simulations. We have made available climate simulations with the Global Ocean Mixed Layer configuration of the Met Office Unified Model, which our research team has performed as part of the BoBBLE project. Our collaborators are using these simulations to assess the role of air-sea feedbacks in the Bay of Bengal on the intra-seasonal variability of the Indian summer monsoon, together with our research team. We bring expertise in coupled climate modelling and the impact of atmosphere-ocean interactions on sub-seasonal variability. Additionally, we have made available single-column simulations of a mixed-layer ocean model, which are designed to understand the roles of particular ocean processes (e.g., the diurnal cycle, horizontal advection, shortwave absorption by chlorophyll) on the structure of the ocean mixed layer in the Bay of Bengal.
Collaborator Contribution Our partners have contributed observations collected during the BoBBLE field campaign in June-July 2016, as well as their expertise in analysing these observations. Our partners have assisted us with the analysis of these observations, including understanding the uncertainty in the individual instruments. Our partners have also created gridded datasets of observations, which we have used an input to our single-column mixed-layer ocean experiments. Our partners also bring many years of expertise in understanding the thermodynamic and dynamic variability of the Bay of Bengal and its interactions with the atmosphere during the Indian summer monsoon.
Impact There have not been any outputs or outcomes yet. This is an ongoing collaboration.
Start Year 2016
Description NCMRWF collaboration for BoBBLE 
Organisation Ministry of Earth Sciences
Country India 
Sector Public 
PI Contribution We have collaborated with the National Centre for Medium Range Weather Forecasting (NCMRWF) to investigate the role of atmospheric forcing and ocean vertical processes on changes in Indian Ocean temperature and salinity during the monsoon season. We have supplied NCMRWF with the multi-column implementation of the K Profile Parameterization (KPP) one-dimensional ocean model, which was developed at the National Centre for Atmospheric Science at the University of Reading. We have assisted NCMRWF with performing simulations in which KPP is driven by reanalysis surface fluxes, including assistance with compiling the model, constructing forcing datasets and interpreting the results.
Collaborator Contribution NCMRWF have performed simulations in which KPP is driven by reanalysis surface fluxes from particularly strong and weak monsoon seasons, in order to understand the role of atmospheric surface forcing and ocean vertical processes on changes in Indian Ocean temperature and salinity in these seasons. NCMRWF have contributed staff time as well as computing resources for performing and analysing these experiments.
Impact There have been no formal outcomes from this collaboration yet, aside from the KPP simulations mentioned above. This is not a multi-disciplinary collaboration.
Start Year 2015