The Panel meeting was preceded by a 2-day Monsoon Symposium (Dec 6-7, 1999), where scientific papers where presented. The first day of the panel meeting was devoted to a Model Intercomparison Workshop, sponsored by the CLIVAR AA-Monsoon Panel in response to the call of the CLIVAR Scientific Steering Group (SSG) for focused efforts to improve our understanding of the causes and consequences of the extensive ENSO disturbances of 1997-1998. Results were presented showing Atmospheric General Circulation Model (AGCM) performance in simulating monsoon circulation, for given Sea Surface Temperatures (SSTs). Further presentations examined the performance of Oceanic General Circulation Models (OGCMs) in simulating the observed SST anomalies (SSTAs), given "observed" surface flux products. The workshop and other talks provided a rather clear picture of the present severe problems facing modelers in their attempts to simulate and predict monsoon activity, using global coupled models.

The CLIVAR SSG also requested the panel to prepare an Implementation Plan for monsoon observations, to be developed over the coming 5 years, with a view towards the possible development of a Monsoon Observing System (MOS) for the following 5 years. It is understood that the development of the MOS will be contingent upon demonstration, from modeling and process studies conducted in the first five years, that the MOS will be valuable for monsoon prediction. During the second and third day (Dec 9 —10), the panel heard presentations on the latest results of monsoon field campaigns and new national and international programs in the AA-monsoon regions, and deliberated on the elements and framework of a 5-10 year CLIVAR AA-Monsoon Implementation Plan. The present document presents an Implementation Plan for an enhanced Monsoon Observing System, based on inputs prepared at various Working Groups held during the last two days of the Panel meeting.

Based on the scientific results presented, the panel emphasized the need to extend the domains of primary interest from the tropical Pacific Ocean, which was the region of major interest in TOGA, to other regions of the tropics and higher latitudes:

".... a compelling reason for this expanded domain is the need to understand and quantify the effect of fluctuations in other major heat sources and sinks of the tropics and subtropics of the global circulation and thereby improve predictions of weather and climate in both local and remote regions .... " (NRC, 1998)

The Asian-Australasian monsoon region is identified as a major system requiring special attention (CLIVAR Initial Implementation Plan, 1998; NRC 1998. The AA-monsoon Panel feels that the field is currently well-poised to make significant scientific advances in understanding the coupled ocean-atmosphere-land system of the monsoon and predicting its variability on time scales of weeks to years.

It is further affirmed that forecasting the variability of the monsoon is a critical issue on a global scale. The annual variability of rainfall and temperature dominate thriving agrarian economic and social structures. The monsoon rains recur each year, but they do so with sufficient variability both in the timing of the onset or within the monsoon season itself to create periods of drought and flood throughout the region. The global context of the monsoons can be seen by noting that in 1990 56% of the global population resided in monsoon regions and 62% in 1995. Thus, besides holding the majority of the population of the planet, the monsoon nations are also the most rapidly growing. Forecasting the intraseasonal and seasonal vagaries of the monsoon, variations that impact agriculture and water resources, are extremely high priority for mankind, and are thus central issues for CLIVAR. The Panel has agreed to undertake the task of writing up an implementation plan, which will serve as a blue print for the next 10 years of international monsoon research. The Implementation plan is written based on the fundamental premise that

There are predictable elements of the Asian-Australian monsoon system, which when fully exploited, will be extremely beneficial to society.

It is hypothesized that predictability of the monsoon lies in the adjacent oceans and land masses, in addition to basin-scale SST fluctuations such as El Niño. For the oceans, SST and the structure of the upper ocean have a degree of predictability in regions where ocean dynamics plays a major role in the surface layer heat budget. For the land, predictability may lie in the snow cover and soil moisture in the monsoon land regions.

 The disastrous droughts and floods in the AA-monsoon region during the great El Niño of 1997-98 have raised important scientific and practical issues regarding monsoon-ENSO relationship, and exposed our lack of understanding in many aspects of monsoon climate variability. In order to achieve the CLIVAR AA-monsoon project’s goal of determining monsoon predictability and to improve monsoon prediction, the Panel recognizes the urgent need to develop a CLIVAR Monsoon Implementation Plan to advance international monsoon research to the next level, and with real benefit to society. Given the significant multi-national efforts already planned for monsoon monitoring, better satellite coverage and improved modeling tools, this is an opportune time for CLIVAR Monsoon Panel to put into action new initiatives for monsoon research.

The Implementation Plan is intended to provide only a broad outline for observational and modeling studies in the monsoon regions, which when properly executed will lead in a few years to establishment of a Monsoon Observing System (MOS). The Monsoon Panel believes that it is more appropriate that the details of a MOS should emerge from further efforts of individuals and groups who are active in monsoon research and observational campaigns in the monsoon regions. In other words, the Monsoon Panel suggests that the same procedure be used as that adopted in the Pacific Ocean in the 1980s and 1990s, where pilot studies defined the requirement and feasibility of TOGA observations, eventually leading to the sustained basin-wide TAO/TRITON array. The Panel believes that a focused effort in modeling and process studies in the monsoon regions over the coming 5 years will lead to the development of a MOS over the next 5 years. The Plan suggests that process studies should each be designed as an entity in itself, to address, with the help of models, specific questions that will lead to better understanding of the physics underlying monsoon predictabi,ity – and hence to improved predictions. Experience in TOGA in developing the Pacific observing system during the TOGA Decade (1985-1994) suggests that a viable design for a long-term MOS should emerge from insights gained from a coordinated set of process studies, together with deliberate design exercises such as Observing System Simulation Experiments (OSSE).

The difficult issue of finding long-term support for the MOS must be addressed by the CLIVAR SSG, and other higher-level bodies. It will require a dedicated effort over several years to publicize the potential value of a MOS, both within the Asian-Australasian region and elsewhere. The first priority here is to ensure the scientific rigor underlying the Implementation Plan. The Panel has no doubt that a properly executed Plan would yield improvements in climate and weather forecasting of substantial benefit to the populations of the monsoon region, and to regions elsewhere influenced by monsoon variability on the 10-year time scale of CLIVAR.

In the 18 months since the last panel meeting and partly as an effort by the Panel to coordinate process experiments and model intercomparison and diagnostic studies of the monsoon anomalies associated with the 1997-98 El Niño, many new insights have been gained based on preliminary modeling and diagnostic studies. A partial list of scientific highlights, described in greater detail in the body of the Implementation Plan, are as follows:

• Monsoon-ENSO relationships are spasmodic and vary on interdecadal time scales. New evidences have suggested that there is a dramatic drop in the correlation between all-India rainfall and the Southern Oscillation Index in the last two decades.
• Intrinsic coupled monsoon rainfall-SST modes have been identified in the western Pacific and the Indian Ocean that contribute significantly to variability of monsoon intensity. Understanding of the physical make-up of these modes is crucial in unraveling fundamental mechanisms of monsoon-ENSO relationships.
• Regarding the existence of a strong biennial variability of the monsoon, a number of theories/hypotheses have been proposed invoking feedback between the land, ocean and atmosphere of the monsoon regime extending from the western Pacific Ocean, the Asian land mass and the Indian Ocean.
• There are new findings suggesting that the Indian Ocean possesses coupled ocean-atmosphere modes that may be independent of ENSO. One of these modes occurs in the form of an oscillating east-west dipole in SST and precipitation. Whereas the relationship to monsoon variability is unclear, it is strongly correlated with the East African monsoon rainfall during September/October/November.
• New evidence suggests that regional components of the monsoon, i.e., the South Asian and the East Asian monsoons, oscillate between active and break phases, in conjunction with the "Madden-Julian oscillations" (MJO) which are modified by the strong north-south pressure gradient and basin-scale SST fluctuations that exist during the established monsoon.
• Interesting and significant results are coming out from recent field experiments e.g., SCSMEX, JASMINE and BOBMEX. For example, preliminary results from jasmine indicates that the ocean-atmosphere heat flux during the established monsoon in the southern Bay of Bengal region is similar to the mean heat flux in the tropical Pacific warm pool. The variability of surface fluxes on intraseasonal time scales in the eastern Indian Ocean is very large, suggesting strong local air-sea interaction. Deliberations at the Hawaii Panel Meeting provided a number of recommendations for action.

With regard to modeling studies, the Panel recommended:

1. Continue to use atmospheric and oceanic gcms in stand-alone mode to understand the sensitivity of the monsoon system (atmospheric and oceanic) to specified boundary conditions. Particular emphasis could be placed on understanding the hydrological and freshwater budgets of the atmosphere and the ocean and, especially, the role of precipitation and river run-off in modulating the upper ocean barrier layer.
2. Continue to conduct intercomparison experiments (in conjunction with Clivar/SMIP, CMIP and amip ii) with special emphasis on the simulation of the ISO and active and break monsoon sequences.
3. Using coupled models and in collaboration with CLIVAR/CMIP, explore the mechanisms and predictability of TBO, and long-term variability of monsoon-ENSO relationships.
4. Noting the very strong diurnal signal that occurred during jasmine in disturbed periods of the monsoon, incorporate the diurnal cycle in ocean, atmosphere and coupled model simulations.
5. Carry out experiments in conjunction with GEWEX, on the effects of land surface processes on monsoon variability, especially as they refer to interannual variations in monsoon forcings.
6. For model validation and diagnostic analysis, utilize and develop high-resolution satellite data sets blending with in-situ observations for precipitation, sea surface temperature and other monsoon parameters; Re-construct, unify and archive various historical data sets currently exist in different monsoon regions

   With regard to process studies, the Panel recommended:

   Consider follow-ups to JASMINE and BOBMEX to study ocean-atmosphere interaction processes in the Bay of Bengal and eastern Indian Ocean, including the Indonesian/northwest Australian regions. These should complement prolonged observations. The Panel therefore warmly welcomes ongoing efforts by Indonesia/US in monitoring the Throughflow; new Indian moorings in the Bay of Bengal, the Arabian Sea and on the equator; planned deployment of Japanese TRITON moorings west of Sumatra; and Australian ARGO/XBT closures northwest of Australia.

7. Consider follow-ups to SCSMEX and studies of the western Pacific warm pool whose variability has been shown to influence variability in the monsoons. This effort should be carried out in conjunction with ongoing and/or planned efforts by Japan, China and Korea in the western Pacific and the marginal seas of East Asia.

• Conduct process studies in the eastern Indian Ocean off Sumatra and the western Indian Ocean to help define an observational strategy first to understand the physical processes associated with the Indian Ocean dipole and then monitor its evolution.
• Continue special land surface observations under Gewex game in asia and the maritime continent.
• Continue studies of the Indonesian Throughflow in order that its evolution over the period of enso events can be measured.
• Maintain a close association and coordination with the gewex cEOps -I program such that process studies are conducted during the ceops -I or extended period.

The Panel believes that the development of a modeling and observing program and its evolution into a mos will be an iterative procedure which needs to be revisited at subsequent annual Panel meetings. In view of the paucity of current long-term observations over the monsoon oceans, and inadequate upper air sounding coverage, the Panel has adopted some initial recommendations which will work toward the establishment of a sustained mos:

• Endorse the current efforts on global monitoring i.e., GCOS, GOOS; encourage the synergistic use of satellite and in-situ observations i.e., geostationary satellites, TRMM, TRMM follow- on, TOPEX, WWW atmospheric sounding network, TAO-array XBT/XCTD lines, etc.
• Embellish the existing xbt/xctd lines that currently exist in the Indian Ocean to obtain long-term monitoring of the entire Indian Ocean north of 30°S using the ship-of-opportunity program.
• Identify critical sites for long term monitoring of the monsoon atmosphere: the Seychelles, Diego Garcia, Maldives, Cocos Island and Port Blair for regular (at least one per day) upper atmosphere soundings.
• Select the Indoex radiation laboratory of Kishadoo (Maldives as the location of an atmospheric wind profiler.
• Encourage the continual deployment of argo floats and drifters in the Indian and the western Pacific Oceans. In the early years, the regions chosen for deployment should be designed to enhance planned process studies.
• Define a strategy for the deployment of an initial atlas/triton buoys in the equatorial Indian Ocean.

 

At each meeting, the Panel will summarize the present status of the entire problem of monsoon prediction, examine and report on the problems encountered in numerical and empirical modeling, and update results of field studies and gauge the potential role of new observing technologies. The Panel will consider plans from different research groups and offer suggestions of how these various studies might be optimized through combination and joint efforts. The Panel sees an important role for the building of bridges between different projects such as gewex, gcos and goos.

In conjunction with GEWEX, the Panel should devise means to tackle the full atmosphere-ocean-land monsoon problem with continued attention on determining the status of monsoon predictability, through process studies and numerical and empirical experimentation. The Panel should be able to offer guidance on a yearly basis towards the goal of developing a mos that would support meaningful and timely monsoon predictions. The international character of the Panel ensures that the stakeholders in monsoon prediction will be fully involved in the development of a mos. Furthermore, a monsoon MOS can do more than help with monsoon prediction. It can help to fulfill the wider aims of GCOS and GOOS. It can also aid in national efforts to forecast tropical cyclones that devastate the region. The Panel will consider these wider aims, as well as the needs to improve monsoon climate prediction, to maximize the usefulness of the MOS to a wide cross-section of the whole community.