2. A MONSOON IMPLEMENTATION PLAN
3.1 Experiments with AGCMs3.2 Experiment with OGCMs
3.3 Interaction with other programmes
4. MONITORING AND SPECIAL OBSERVING SYSTEMS
4.1 Problems related to moisture and cloudiness4.2 Problems in understanding SST anomalies
4.3 Land surface processes
5. DATA REQUIREMENTS AND EMPIRICAL STUDIES
6. PROCESS STUDIES AND FIELD CAMPAIGNS
1. INTRODUCTION The first session of the CLIVAR Monsoon Panel was held in Panaji, Goa, India, from 19-22 November 1996, overlapping with the 5th session of the TAO Implementation Panel (TIP). The Monsoon Panel session was formally opened by Dr M. R. Nayak, who warmly welcomed the participants on behalf of Dr Desa, Director of the Indian National Institute of Oceanography (NIO), the local host for the two meetings. In reply Dr M. Coughlan, Director of the International CLIVAR Project Office (ICPO), thanked Dr Nayak and his colleagues for their preparatory work for the meetings and for organising the very comfortable venue and efficient working arrangements at the Hotel Mandovi. While a number of members of the Panel were unable to attend due to other commitments and travel constraints, the combination of the two meetings helped make up for some of this loss of expertise. The work of the Monsoon Panel session began with an overview of the current status of CLIVAR by Dr Coughlan who stressed the urgency for outlining a CLIVAR initiative on the Asian-Australian Monsoon. The ICPO had been tasked by the Joint Scientific Committee of the World Climate Research Programme with preparing a first draft of the CLIVAR Implementation Plan for its 18th Session in Toronto in March, 1997. Dr Coughlan also noted that CLIVAR, with its legacy deeply rooted in TOGA and WOCE, had enormous potential to bring new dimensions to the existing national and regional activities on the monsoon. Chiefly these were the capacity for developing a global perspective on monsoons and the experience to incorporate a coherent ocean component in to the study of monsoon structures and variability. The co-chairs Dr Bill Lau and Prof. Akimasa Sumi reviewed the terms of reference of the Panel which are as follows:
In co-operation with CLIVAR NEG-1 and other CLIVAR Panels, the CLIVAR Monsoon Panel will:
It was agreed that these terms of reference were appropriate to the work of the Panel and no amendments were necessary at this time. Nevertheless, the task of addressing the monsoon as set down above was very broad, and it would be necessary for the Panel to develop a set of priorities that achieved maximum return for the resources that were likely to be available. To assist in defining an overall structure of monsoon climatology, Dr Lau described the monsoon being comprised of 'fast' atmospheric and 'slow' oceanic/land systems that interacted at various intermediate time-scales through processes that coupled the atmosphere with the land and with the adjacent oceans (Figure 1). Many of the phenomena associated with the 'fast/slow' system are manifested in the form of 'intra-seasonal oscillations' (ISOs), such as the 30-60 day Madden-Julian Oscillation (MJO), a quasi periodic set of waves of disturbed weather that can be tracked moving eastward in low latitudes from the Indo-Pacific region, through the longitudes of the maritime continent, and into the Pacific Ocean. The character of a summer monsoon in any given year could generally be characterised as 'strong' or 'weak' depending on the relative frequency of active and break periods associated with the passage of intra-seasonal disturbances. Following this opening session, the Monsoon Panel and TIP combined for a series of scientific presentations on current monsoon and low latitude climatology.
Back to the table of contents
2. A MONSOON IMPLEMENTATION PLAN To begin the process of preparing a draft plan for a CLIVAR Asian/Australian monsoon initiative, the Panel addressed itself to defining the priority science issues and to developing a commensurate set of goals and objectives. Following considerable discussion, these were distilled to the following. Goals The determination of: the limits of predictability of the Asian-Australian-African monsoon in the climate system, the relative contribution of the slowly varying boundary conditions and the internal dynamics, and the impact of the monsoon on the predictability of the global climate system. To achieve these goals, the CLIVAR Monsoon Initiative, would need to achieve the following: Objectives
In view of the importance of the ISOs to monsoon climatology, it was agreed that a study of them should form a key component of the first phase of the CLIVAR initiative. A related aspect that would also require the early attention of the CLIVAR initiative is the determination of any potential skill in predicting the annual onset of a wet monsoon. While the overall list of objectives is quite daunting, the Panel was reminded of the many existing efforts of a national and regional nature that could be linked to and made an integral part of the CLIVAR initiative. These national and regional programmes are listed and outlined in Appendix C. To address the objectives, four key areas of activity were identified. Sub-groups of Panel members were formed and charged with preparing initial texts. The four areas and sub-groups are: Modelling and Prediction (Shukla, Goswami), Monitoring and Special Observing Systems (Godfrey, Sumi, Pu), Data Requirements and Empirical Studies (McBride, Li), Process Studies and Field Campaigns (Kang, Yasunari) The outcome of this exercise is summarised here; the fuller text emanating from the deliberations of the Panel will be embodied in the draft CLIVAR Implementation Plan. Back to the table of contents
3. MODELLING AND PREDICTION The modelling and prediction programme recommended by the panel aims at achieving the following primary goals, namely:
Although there exists considerable empirical evidence that monsoon variability is related to several slowly varying boundary conditions (e.g. SST, snow cover, soil wetness etc.), these relationships seem to explain only about 25% of the variability. Therefore, while there is potential predictability of the monsoon, it is yet to be determined what part of this predictability can be realised. One of the important questions that needs to be investigated is whether the character of the intra-seasonal variability is determined by boundary conditions or is independent of the boundary forcing. The present status is that several atmospheric general circulation models indicate high potential predictability of the monsoon. However, the actual skill of these models for prediction of quantity like the Indian rainfall is still small (anomaly correlation). This limited skill is due likely to a combination of two reasons: a) an inability of the climate models to simulate mean monsoon climate realistically; and
b) the contribution of internal dynamics It is encouraging that several high resolution AGCMs and regional climate models nested in AGCMs have shown promise in realistically simulating the regional features of the monsoon. Such model developments, which will ultimately lead to better simulations of the mean monsoon, are recommended. However, the effort is partly that which is required to improve the parameterizations of physical processes. Accordingly, the model development activity needs to be closely co-ordinated with the plans for Process Studies discussed below. Simple and intermediate model studies are also recommended to derive deeper understanding of the physical processes. As monsoon climate is certainly a coupled ocean-atmosphere-land phenomenon, ideally coupled GCMs should be used to address the monsoon predictability. While tremendous progress has been made in coupled modelling related to ENSO involving the Pacific ocean, coupled modelling efforts over Indian monsoon region has hardly begun. This is partly due to the fact that uncoupled results of how ocean SST influences the LF variability of the atmosphere and how atmospheric winds and heat fluxes influence the LF variability of the ocean have not been clearly established in this region. In view of this status there is an urgent need to assess the ability of the new generation of AGCMs to simulate the seasonal mean, interannual variability and intra-seasonal variability of the monsoon and to make quantitative estimate of the predictability (contributions from slowly varying boundary conditions versus internal dynamics). Keeping this background in mind, the panel recommends a series of six numerical experiments to achieve the objectives listed above. Experiments with AGCMs
Experiments with OGCMs
3.1 Experiments with AGCMs
This modelling activity aims to achieve the following objective
a) Reanalysis products
b) In situ observations in the region, Station data as well as Satellite data In this connection, Indian data on daily precipitation, circulation and cloudiness data from INSAT at about 1ºx 1º resolution will be required. The earlier AMIP simulations indicated that most AGCMs had serious problems in simulating the regional features such as the Indian precipitation. During the last five years, however, it has been demonstrated that many AGCMs (e.g. JMA T106 Version, ARPEGE, COLA etc.) and regional nested climate models (COLA: Vernekar et al, 1996 and Hadley Centre: Bhaskaran et al, 1996) are beginning to simulate the seasonal mean precipitation close to observations. This is why it is opportune to carry out a similar set of experiments with the new models. The panel recommended that a recent 15 year period from 1979 to 1994 be taken that includes the strong 82/83 and 1987/88 ENSO events together with the persistent El Niño-like conditions during 1992 and 1994. It is suggested that for validating the simulation of the monsoon, several monsoon indices be adopted, based on precipitation (e.g. core region) as well as circulation (e.g. Webster and Yang, 1992 or Goswami et al.,1996).
The objective of these sets of experiments is to assess the limit on predictability of the monsoon and to make quantitative estimate of contributions from slowly varying boundary forcings and internal dynamics. For this purpose two sets of experiments are recommended: For the ensemble AMIP runs, the same period as in (i) could be prescribed, which would give one of the ensemble already. The summer seasons for the same years might be considered for seasonal prediction as well.
Some evidence is beginning to emerge that the intra-seasonal oscillations of the monsoon play an important role in determining the seasonal mean and interannual variability. There is also some evidence of similar intra-seasonal oscillations in the ocean. However, it is not clear whether coupling with the ocean is essential for the atmospheric ISO or that they can be sustained entirely by the atmosphere. It is also possible that the land processes play a more important role than the oceanic processes. To establish the relative roles of the oceans and the land, the following AGCM experiments are recommended.
These experiments would be carried out with a few AGCMs exhibiting reasonable monsoon climatologies.
The amplitude of interannual variations of the SST over the Indian ocean during the boreal summer is rather small. However, these variations are on high mean temperatures and as such can still influence the large scale circulation in a significant way. Conceptually, SST anomalies over different regions over the Indian ocean could influence the monsoon through different mechanisms. For example, positive anomalies in the south Indian ocean and Arabian sea are expected to enhance continental precipitation through enhanced moisture flux. However, a positive SST anomaly over the equatorial Indian ocean may cause decrease in continental precipitation by favouring the oceanic ITCZ over the continental one. Similarly, a positive SST anomaly over the Bay of Bengal is expected to enhance continental precipitation through increased synoptic activity. These scenarios need to be addressed using AGCMs with realistic mean monsoon climatologies and it is recommended that such a set of experiments be conducted by a few groups. The panel notes that CLIVAR NEG-1 has also proposed a series of experiments with AGCMs to understand the dynamics of interannual variability of the monsoon. The panel recognises that those experiments are also important and to avoid duplication, the modelling activities proposed by the Monsoon Panel will be closely co-ordinated with NEG-1. 3.2 Experiments with OGCMs As mentioned in the preamble, the questions related to the uncoupled mode in the monsoon region have not yet been addressed adequately in order to proceed with confidence to address questions related to coupled modelling. The present OGCMs have SST simulation errors of the order of 2ºC or more. This is due to two major deficiencies. First, most OGCMs have a significant thermocline drift in the monsoon region; the other is related to uncertainties in the estimation of the net heat flux. As the mixed layer dynamics is important in this region, a good estimate of the heat flux is essential. A layered model with an atmospheric mixed layer model seems to show promise (Murtugudde et al, 1996). As with AGCMs, these model developments should go on. To accelerate the process of these model developments and to get some answer to the uncoupled questions mentioned above, the panel recommends the following two sets of experiments.
These will be AMIP-type runs. Here a number of OGCMs will be forced by the same set of wind stress and heat flux products. These intercomparison exercises would not only establish the status of the OGCMs, but they would also help identify possible causes of the problems. For validation and verification of the OGCM simulation, circulation data acquired by NIO (India) may be useful. 3.3 Interaction with other programmes As discussed earlier, the model development efforts will closely interact with process studies group. Similarly, the data requirement group (see below) has recommended that high resolution (daily rainfall and INSAT OLR) regional data should be obtained for studying the variability of monsoon rainfall as well as for model verification. Once such data becomes available it is recommended that a "regional reanalysis" project be launched that will use a good regional climate model nested in an AGCM. This will help establish a data set for detailed three-dimensional studies of various processes related to monsoon variability. As more reliable coupled GCMs become available over this region, at a later stage of the programme, two additional sets of numerical experiments might be considered.
Here a control run would first be run. Then the monsoon would be artificially perturbed (e.g. by dumping a lot of snow over Eurasia or by drying the land) and the effects on the evolution of ENSO examined. Similarly, ENSO could be artificially perturbed and the effects on the monsoon examined.
Back to the table of contents
4. MONITORING AND SPECIAL OBSERVING SYSTEMS 4.1 Problems related to moisture and cloudiness A major constraint on understanding monsoon variability is the lack of adequate data, particularly over the surrounding ocean regions, but also over land. Satellites in principle offer a solution but typically, the moist and cloudy nature of the monsoon atmosphere introduces severe difficulties, e.g.:
The effort to improve the situation is being addressed elsewhere in the research community, particularly in GEWEX. The launch of TRMM and its associated validation efforts should improve the rainfall estimates. New, lower-cost technologies are being also developed, such as autonomous atmospheric sounders (e.g. the Aeorosonde), improved rain gauges for use on ocean moorings, and the use of commercial aircraft for soundings and flight-level sampling. CLIVAR needs to work closely with the relevant GEWEX scientists to ensure that creative solutions, which are not only specific to measurement problems in Asian-Australian monsoon region, can be found. 4.2 Problems in understanding SST anomalies A second serious difficulty for monsoon research is that outside the equatorial Pacific our understanding of the mechanisms of SST anomalies is quite limited. In the northern Indian Ocean these mechanisms are complex and unique. Ocean modelling efforts to improve our understanding are underway, using present estimates of surface fluxes, but as described elsewhere in this Plan, all relevant surface fluxes contain serious errors. Furthermore, the subsurface data in the monsoon oceans is particularly limited. Various proposals for enhancement of subsurface data are being considered, by enhancing the existing XBT network; placement of moored and drifting buoys; and Profiling ALACE floats should be encouraged or proposed under CLIVAR-sponsored research projects. Observing System Sensitivity Experiments (OSSEs) also can be conducted to assess the best use of such resources for improving understanding of SST anomalies patterns of relevance to the monsoons. Particular attention needs to be placed on the determining the optimum deployment of the presently high-cost salinity sensors. 4.3 Land surface processes Finally, in order to understand the predictability of the monsoon and the mechanism of boundary forcing and internal dynamics, the continuous monitoring of land surface processes is necessary, especially those associated with controlling the availability of soil moisture.
Back to the table of contents
5. DATA REQUIREMENTS AND EMPIRICAL STUDIES The list of data identified by the sub-group initiative for a CLIVAR AA-Monsoon initiative includes the following: Basic Observations
Derived Analyses
The list of possible empirical studies is virtually open ended, however, several studies could be initiated early to reveal some important characteristics of monsoon variability, for example, along the following lines:
To resolve some of the problems experienced by researchers in accessing data from the monsoon regions, the subgroup identified the following key issues:
and proposed the following practical measures:
Back to the table of contents
6. PROCESS STUDIES AND FIELD CAMPAIGNS In addition to the need for the CLIVAR Monsoon initiative to work closely with the various national and regional activities, the Panel recognised the value in establishing co-operative links with other parts of CLIVAR and with other relevant parts of the WCRP, especially GEWEX. Further, the relevance of several projects proposed by the CLIVAR NEGs and the Upper Ocean Panel to the monsoon initiative are also very important; the ongoing research being conducted on the emerging TOGA-COARE data sets, e.g. by the TOGA COARE Flux Group, appears especially relevant. These data sets, collected in the Western Pacific Warm Pool (WP), between November 1992 and February 1993, during the international TOGA-COARE experiment are a particularly rich legacy from TOGA to CLIVAR. The data are of unprecedented breadth and accuracy, and provide a massive resource for examining fundamental questions, such as:
By analogy with the earlier GATE (GARP Atlantic Tropical Experiment), the most important results of COARE, in terms of rigorously validated algorithms for climate models, are expected to appear about a decade after the field phase of the experiment. It is therefore essential that CLIVAR continues to support modelling and data analysis activities associated with COARE for the first five years of CLIVAR, and to foster interaction between data analysts and modellers. There are several key areas within the 'fast' components of the climate system where GEWEX will be a critical contributor to a CLIVAR monsoon initiative, notably for cloud processes (ISCCP and GCSS), precipitation (GPCP and TRMM), and water vapour (GVaP). Particular attention was drawn to the cumulus parameterization problem and the problems of the monitoring and measurement of spatial and temporal variations in atmospheric water vapour over the tropical oceans. While these elements may not be the focus of the monsoon plan, it is critical that they receive the attention necessary for the success of the CLIVAR initiative. The Panel will therefore need to establish appropriate dialogues with its GEWEX counterparts to ensure the establishment of collaborative frameworks and activities. The existence of the series of regional projects being conducted under the umbrella of the GEWEX Asian Monsoon Experiment (GAME) will provide a number of potentially fruitful opportunities for collaboration on land surface processes. The following specific possibilities for study were presented for consideration by the Panel for inclusion in the CLIVAR Implementation Plan.
With respect to the study of these two mechanisms, rich sources of observed data are becoming available in the western Pacific and East Asian region due to various field experiment programmes such as TOGA-COARE, TAO, GAME, SCSMEX, and KORMEX (Korea Monsoon Experiment). In contrast, the Indian subcontinent and surrounding seas have not had the same concentration of systematic observations or programmes of field campaigns as the areas further to the east. The Indian Ocean Experiment (INDOEX) offers an opportunity to redress some of this imbalance. INDOEX is being planned to investigate the large-scale effects of aerosol and other emissions from Asia on the composition of the atmosphere over the Indian Ocean. In the Indian dry season (Spring), extensive convective clouds form over the oceanic "warm pool" and the juxtaposition of very clean air from the Southern Hemisphere with the emissions from a population of one billion makes for an ideal natural laboratory to study the interaction of chemistry, microphysics, radiation, and climate. In these respects then, INDOEX is a highly valuable opportunity for providing a reliable, enhanced multi-purpose data set for the region. The field programme for INDOEX is scheduled for 1998-1999 and the Panel believed this offers distinct possibilities for collaborative efforts. To explore the various options for collaboration the Panel agreed that Dr Shukla be asked to form a small task team with representation from national CLIVAR interests and the proposed field experiment sub-groups with which CLIVAR could usefully join forces. In addition to collaborating in the field experiment planning underway for INDOEX, CLIVAR could begin feasibility studies of the following experiments:
The reanalysis data sets from NCEP, ECMWF and NASA will be valuable for the large scale process studies, and monsoon researchers should make maximum use of these readily available data sets.
TOGA-COARE data should be fully utilised for the study of tropical processes, and for the planning of ongoing experiments such as GAME, TIMEX, KORMEX. It is expected that these experiments will also produce valuable data sets for research in these areas and will have the added benefit of extending the domain to include extra-tropical processes.
Back to the table of contents
Last modified: Tue May 6 04:37:54 PDT
