Atmospheric Modeling & Research


Data Integration and Computing Facilities: A complete end-to-end forecasting system that includes acquisition of data from various observing systems, linkage to a central data processing system, their utilization in the numerical models, providing a state-of-the-art IT based environment to all forecasters across the country has been installed. This involves integration of all observations and overlaying them on model outputs and synoptic charts along with proper visualization and customized dissemination of weather forecast to the end users.

During the first phase of the modernization of IMD, along with the commissioning of state-of-the-art observing systems, application module “SYNERGIEe” has been installed at 34 forecasting offices all over India. These have the capability to visualize multiple layers of observation and forecast overlayed on each other, thus providing to the forecaster the capability to assimilate terabytes of information before issuing weather forecast. A Public Weather Service System(PWS) ‘METEOFACTORY ’ enables the forecaster to generate required customized presentation for the automatic delivery of products to the visual and print media viz. newspaper, TV, farmers, shipping, etc. The Central Information Processing System (CIPS) with a super-computer at the back-end has given a paradigm shift from Observation Network Management towards integrated information processing systems providing fully customized services to the user community. It is equipped with state-of-the-art hardware and software to handle, manage, store, process and archive all data and forecast products used in operational weather forecasting. It consists of seven sub-systems comprising data acquisition, data policy management, data centre, tpolicy management, task centre, backup policy management . The CIPS Data Centre provides flexibility to ingest any kind of meteorological data like satellite, radar, etc. with regard to the assessment of current weather and prognosis for the future.

A set of four High Performance Computing systems (HPCS), one each at INCOIS, IITM, IMD, and NCMRWF have been installed for global data processing, assimilation and numerical weather prediction (NWP) for weather and ocean state forecasting services. The combined strength of HPCS in ESSO is about 125 TeraFlops, which has significantly improved coupled atmospheric–ocean modelling capability. High End Servers at 10 centres namely, RMC Delhi, Kolkata, Chennai, Mumbai, MC Chandigarh, Ahmedabad, Nagpur, Bhubaneswar, Bangalore and Hyderabad were commissioned.

Weather Modeling Framework: With the commissioning of HPCs, Global Forecast System (GFS T382/L64) having spatial resolution of 35 km was made operational at IMD, incorporating Grid Point Statistical Interpolation (GSI) scheme for global data assimilation for the generation of global scale forecasts up to 7 days in advance. At NCMRWF, continuous efforts are on to enhance the accuracy and reliability of the forecasts by increasing resolution, improving physical processes, data assimilation, optimizing use of satellite and DWR data , coupling land, ocean and atmosphere, ensemble forecasting along with extensive verification and validation. The horizontal resolution of the Global Forecast System (GFS) was increased from T254L64 (50km) to T382L64 (~35km) along with assimilation of direct satellite radiances which was subsequently made operational at IMD. Experimental runs with higher resolution (T574L64(22km) GFS and unified model (N512L70(25Km) with 4D VAR assimilation) models are underway. It is planned to make the high resolution system operational after extensive evaluation and verification studies.

Meso-scale forecast system WRF (ARW) with 3DVAR data assimilation is being operated daily twice, at 27 km, 9 km and 3 km horizontal resolutions for the forecast up to 3 days in advance using initial and boundary conditions from the GFS T382/L64 system. At regional centres as mentioned above, very-high-resolution mesoscale models (WRF at 3 km resolution) are made operational to generate high resolution regional scale forecasts. Typical customization of NWP-based forecast products is carried out in support improving cyclone warning service. Doppler weather and mesoscale WRF-model-based Nowcast system was exclusively for the national Capital of Delhi to generate site specific hourly scale weather forecasts. Global model assimilation utilizing 4D VAR is being done on experimental basis for WRF model for generating regional scale data assimilation.

In an effort o take care of the uncertainties in the initial conditions, model dynamics and model physics, Multi Model Ensemble (MME) forecasting project for providing rainfall forecasts during the monsoon season has been developed. Rainfall forecast data from 4 global models of India and outside India, i.e. NCMRWF, NCEP,USA, UKMO,UK and JMA, Japan for the summer monsoon period were used to train the MME algorithm coefficients. MME forecasts of rainfall in short and medium range were made available for operational real time use for the Indian region. District Level Quantitative five-days weather forecasts based on MME system are being generated in support of rendering district scale Agro-Meteorological Advisory Service (AAS) of India during the current five year plan.

Polar WRF is implemented to provide day-to-day short range (48 hours) weather forecast for the Maitri region in the Antarctica. All these NWP products are routinely made available on the IMD web site ( in support of Antarctic Expedition. NCMRWF was regularly providing NWP guidance from its global and regional analysis and forecast products in support of first ever South Pole Expedition.

A new state-of-the-art Climate Forecast System (CFS) based on coupled ocean-land-atmosphere model, was setup at IITM and numerous sensitivity experiments were carried out to test the model skill in predicting the Indian Summer monsoon rainfall on seasonal time Further, the model horizontal resolution is increased from T62 (~ 200km) to T126 (~ 110 km) that resulted in better reproduction of spatial distribution of monsoon rainfall over India. This effort led to the utilization of coupled model framework to provide experimental forecast assessment of the Indian Summer monsoon seasonal mean rainfall and its distribution since March 2010. Similarly, prediction of active/break spells of the Indian summer monsoon is implemented in IMD through collaboration with IITM

The Numerical Weather Prediction (NWP) products from mesoscale model WRF were supplied to BARC at 9km resolution for the four nuclear sites namely Trombay, Kalpakkam, Kaiga and Narorafor their Indian Real time Online Decision Support System for Offsite Nuclear Emergency (IRODOS).

Last Updated On 04/09/2015 - 12:37
Back to Top