"The Meeting of the Advisory Group on enhancing the forecasting capabilities forNorth Indian Ocean Storm Surges (IIT-D storm surge model upgrade) was held in India atIndian Institute of Technology Delhi (IIT Delhi) during 14-17 July 2009, under the cochairmanshipof Dr Boram Lee (IOC of UNESCO) and Dr Val Swail (JCOMM). Theinternational experts reviewed current status/performance of operational storm surgeforecasting model (IIT-D Model) in the North Indian Ocean (NIO) region and addressedrequirements for upgrading and improving model performance, considering the needs of theregion following the recommendations made at the first JCOMM Scientific and TechnicalSymposium on Storm Surges held during 2-6 October 2007 at Seoul, Korea Republic(JCOMM2007SSS, http://www.surgesymposium.org). Experts also deliberated on setting upthe medium-term and long-term technical workplan for the project on improving storm surgeforecasting capabilities.Highlights and recommendations from the workshop are the following:While the storm surge prediction capability of India and further the North IndianOcean region was generally satisfactory, the forecast was presently given only at the districtlevel - the immediate goal was to provide forecasting with finer spatial resolution (smallgeographical regions) with better predictability. Experts agreed that, to be able to do this, thenumerical models as well as the observational networks have to be improved. Additionalresearch needs to be done on how to incorporate into operational use, some other importantprocesses that presently were not included in the forecast models. With these goals in mind,the workshop has highlighted the requirements under observational data, categorized intometeorological, boundary, hydrological and location specific.The experts group stressed the importance of the storm surge forecasting inprobabilistic terms, somewhat akin to the routine weather forecasts at present. This is a morerealistic approach, since it is scientifically unsound to issue definitive diagnostic forecasts,when there is so much uncertainty in the input data to the numerical models. It has beenshown clearly that tide-surge interaction could change the value of the Total Water LevelEnvelope (TWLE) as well as the timing of the occurrence of the peak surge. Hence it isimportant the numerical models include the non-linear interactions between tide and stormsurge. At present the only meteorological data input to the storm surge models of IIT Delhiare the pressure drop and the radius of maximum winds, in addition to the cyclone track.Since observed wind fields are now becoming more and more routinely available in real time,the numerical models should incorporate data assimilation techniques such as Kalman Filterto include observed wind data. At present the wind wave model is not part of the storm surgemodel. An operational wind wave model should be dynamically coupled to the storm surgemodel. For the cyclone season of 2010 and beyond, the coupled model (surge plus tide pluswind wave) should be first tested in a hindcast mode, and once it is established that themodel performance is satisfactory. it can then be used operationally.One of the key recommendations on data was on bathymetric data. 100 mhorizontal resolution over the continental shelf is recommended in order to enable highresolution modelling. This data set should be updated every five years over the shelf, andevery three years in the river delta regions such as the Meghna, Hooghly, Godavari, Krishna,Ayeyarawady etc. Another important requirement was to have an optimum network of tidegauges in the vulnerable coastal areas with a temporal resolution of one minute averagedsea level data. At present the existing and the planned tidal network together is generallysatisfactory for India, with a wish list of a few more gauges located at the head of the Bay ofBengal. It was recommended to install at least two more gauges in Bangladesh and threemore gauges in Myanmar, for the storm surge modelling and monitoring purpose.The two most important meteorological parameters were again emphasized by theExperts to be more accurately monitored during the events, in addition to location of thecyclone landfall - the pressure drop and the radius of maximum sustained winds. Aircraftreconnaissance and dropsonde data may enhance the quality of the input data.On the Hydrological side, we need data on area-weighted rainfall in rivercatchments as well as station data. In addition to this, for Hydrological input to surge models,we need to enhance the existing river gauge network for data on river runoff.A most useful product will be storm surge data dossiers, which include all data frompost-event surveys and also detailed data on road network, escape routes, nearby highground (hills and mountains) and coastal infrastructure.For the computation of inundation, we need data on shore topography, with aresolution of 5 m in the horizontal and 0.5 m in the vertical, with updates done preferablyevery decade. The use of crest gauges as well as chemically treated ribbons for mappingthe extent of horizontal inundation and depth should be considered.For improvements to cyclone track prediction, the Multi Model Ensemble (MME)should be enhanced through improvements in the meso-scale Numerical Weather Prediction(NWP) model, and nowcasting assimilation of the remote sensing (satellite and radar) and insitu(buoy and ship) based data.With regard to the capacity development and outreach, it was suggested to continuetraining on IIT-D model development and operation, in collaboration with ongoingprogrammes by WMO and IOC.Based on advisory input as described above, the 3-year workplan was agreed bythe Expert group and IIT Delhi expert team, to improve the predictability of the IIT-D StormSurge model. Priorities were set up and agreed in view of enhancing storm surgepredictability through the IIT-D Storm Surge Model, taking into account; 1) ongoing plans andactivities in the North Indian Ocean region; 2) timeline and workplan for IIT model upgrade,and; 3) feasibility of each recommendation." |
