Search

The rainfall forecast of global forecasting system (GFS) model running at the India Meteorological Department (IMD), i.e., GFS T574 (25 km) and GFS T1534 (12.5 km) is verified at its natural resolution. A verification study is conducted for the rainfall forecast over Indian window during the summer monsoon season (June–September) 2016. Regional verification of rainfall forecast was also carried out over five homogeneous regions of India, i.e., north India, west coast of India, northeast India, central India and peninsular India. The skill of the model forecast is also evaluated for yes/no, categorical and few heavy rainfall cases over India. Results show that both T1534 and T574 models are able to provide more realistic spatial distribution of seasonal mean rainfall over India. This study also shows that both the model forecasts are skillful in capturing the 24 h accumulated rainfall over climatologically heavy rainfall regions, but the magnitude and location fluctuated. The rainfall forecast for different rainfall categories in terms of different statistical skill scores, i.e., ratio score, HK, bias, hit rate and HKQ showed a relatively high skill for T1534 compared to T574 over most parts of the country from the day 1 to day 5 forecasts. It is also found from this study that the skill in predicting day 1 to day 3 forecasts of both the models was found to be good and usable for all parts of the country except for high terrain regions in India. The results documented here are expected to be useful to the operational forecasters in day-to-day weather forecasting over Indian monsoon regions.

During August 2018 and 2019 the southern state of India, Kerala received unprecedented heavy rainfall which led to widespread flooding. We aim to characterize the convective nature of these events and the large-scale atmospheric forcing, while exploring their predictability by three state of the art global prediction systems, the National Centre for Environmental Prediction (NCEP) based India Meteorological Department (IMD) operational Global Forecast System (GFS), the European Centre for Medium Range Weather Forecast (ECMWF) integrated forecast system (IFS) and the Unified Model based NCUM being run at the National Centre for Medium Range Weather Forecasting (NCMRWF).Satellite, radar and lightning observations suggest that these rain events were dominated by cumulus congestus and shallow convection with strong zonal flow leading to orographically enhanced rainfall over the Ghats mountain range, sporadic deep convection was also present during the 2019 event. A moisture budget analyses using the ERA5 (ECMWF Reanalyses version 5) reanalyses and forecast output revealed significantly increased moisture convergence below 800 hPa during the main rain events compared to August climatology. The total column integrated precipitable water tendency, however is found to be small throughout the month of August, indicating a balance between moisture convergence and drying by precipitation. By applying a Rossby wave filter to the rainfall anomalies it is shown that the large-scale moisture convergence is associated with westward propagating barotropic Rossby waves over Kerala, leading to increased predictability of these events, especially for 2019.Evaluation of the deterministic and ensemble rainfall predictions revealed systematic rainfall differences over the Ghats mountains and the coastline. The ensemble predictions were more skilful than the deterministic forecasts, as they were able to predict rainfall anomalies (>3 standard deviations from climatology) beyond day 5 for August 2019 and up to day 3 for 2018.

A global forecast system model at a horizontal resolution of T1534 ( $${\sim }12.5\, \hbox {km}$$ ) has been evaluated for the monsoon seasons of 2016 and 2017 over the Indian region. It is for the first time that such a high-resolution global model is being run operationally for monsoon weather forecast. A detailed validation of the model therefore is essential. The validation of mean monsoon rainfall for the season and individual months indicates a tendency for wet bias over the land region in all the forecast lead time. The probability distribution of forecast rainfall shows an overestimation (underestimation) of rainfall for the lighter (heavy) categories. However, the probability distribution functions of moderate rainfall categories are found to be reasonable. The model shows fidelity in capturing the extremely heavy rainfall categories with shorter lead times. The model reasonably predicts the large-scale parameters associated with the Indian summer monsoon, particularly, the vertical profile of the moisture. The diurnal rainfall variability forecasts in all lead times show certain biases over different land and oceanic regions and, particularly, over the north–west Indian region. Although the model has a reasonable fidelity in capturing the spatio-temporal variability of the monsoon rain, further development is needed to enhance the skill of forecast of a higher rain rate with a longer lead time.

The forecast for 655 districts and 6500 blocks had been prepared and implemented on 1st June, 2014. The procedure for getting forecast for the districts and blocks in India including altitude corrections is based upon regular (0.25 × 0.25) grid output from the T-574 Model and output from 9 km WRF model. A verification study for rainfall forecast at 0.25 × 0.25 degree grid for Indian Window (0-40° N and 60-100° N) is also conducted, which had indicated that skill of the rainfall forecast is good for all parts of the country except oceanic islands and high terrain regions and one can down scale to any level, down to the blocks, the skill scores will not differ much. A detailed verification study for the skill of the forecast at block level for all the eight weather parameters for which the forecast was issued is conducted. The skill of the rainfall forecast is obtained for categorical forecast and as well as for yes/no forecast. The skill scores for rainfall had indicated that highest value of Hanssen and Kuiper (HK) score is 0.44, Hanssen and Kuiper score for quantitative rainfall (HKQ) is 0.18, Ratio score for yes/no forecast is 90 percent and Hit rate (HR) is 0.83. The detailed verification study for the block level weather forecast for monsoon 2014 is presented in the paper and the skill found is good. The study indicates that model forecast has the potential to be used for the block level forecast after making the quick value additions for which hints are given in the conclusion part.

The output from Global Forecasting System (GFS) T574L64 operational at India Meteorological Department (IMD), New Delhi is used for obtaining location specific quantitative forecast of maximum and minimum temperatures over India in the medium range time scale. In this study, a statistical bias correction algorithm has been introduced to reduce the systematic bias in the 24–120 hour GFS model location specific forecast of maximum and minimum temperatures for 98 selected synoptic stations, representing different geographical regions of India. The statistical bias correction algorithm used for minimizing the bias of the next forecast is Decaying Weighted Mean (DWM), as it is suitable for small samples. The main objective of this study is to evaluate the skill of Direct Model Output (DMO) and Bias Corrected (BC) GFS for location specific forecast of maximum and minimum temperatures over India. The performance skill of 24–120 hour DMO and BC forecast of GFS model is evaluated for all the 98 synoptic stations during summer (May-August 2012) and winter (November 2012–February 2013) seasons using different statistical evaluation skill measures. The magnitude of Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE) for BC GFS forecast is lower than DMO during both summer and winter seasons. The BC GFS forecasts have higher skill score as compared to GFS DMO over most of the stations in all day-1 to day-5 forecasts during both summer and winter seasons. It is concluded from the study that the skill of GFS statistical BC forecast improves over the GFS DMO remarkably and hence can be used as an operational weather forecasting system for location specific forecast over India.

Performance of national centers for environmental prediction based global forecast system (GFS) T574/L64 and GFS T382/L64 over Indian region has been evaluated for the summer monsoon season of 2011. The real-time model outputs are generated daily at India Meteorological Department, New Delhi for the forecasts up to 7 days. Verification of rainfall forecasts has been carried out against observed rainfall analysis. Performance of the model is also examined in terms of lower tropospheric wind circulation, vertical structure of specific humidity and precipitable water content. Case study of a monsoon depression is also illustrated. Results obtained show that, in general, both the GFS T382 and T574 forecasts are skillful to capture climatologically heavy rainfall regions. However, the accuracy in prediction of location and magnitude of rainfall fluctuates considerably. The verification results, at the spatial scale of 50 km resolution, in a regional spatial scale and country as a whole, in terms of continuous skill score, time series and categorical statistics, have demonstrated superiority of GFS T574 against T382 over Indian region. Both the model shows bias of lower tropospheric drying and upper tropospheric moistening. A bias of anti-cyclonic circulation in the lower tropospheric level lay over the central India, where rainfall as well as precipitable water content shows negative bias. Considerable differences between GFS T574 and T382 are noticed in the structure of model bias in terms of lower tropospheric wind circulation, vertical structure of specific humidity and precipitable water contents. The magnitude of error for these parameters increases with forecast lead time in both GFS T574 and T382. The results documented are expected to be useful to the forecasters, monsoon researchers and modeling community.

India Meteorological Department has implemented district level medium range rainfall forecast system applying multimodel ensemble technique, making use of model outputs of state-of-the-art global models from the five leading global NWP centres. The pre-assigned grid point weights on the basis of anomaly correlation coefficients (CC) between the observed values and forecast values are determined for each constituent model at the resolution of 0.25 ◦ ×0.25 ◦ utilizing two season datasets (1 June–30 September, 2007 and 2008) and the multimodel ensemble forecasts (day-1 to day-5 forecasts) are generated at the same resolution on a real-time basis. The ensemble forecast fields are then used to prepare forecasts for each district, taking the average value of all grid points falling in a particular district. In this paper, we describe the development strategy of the technique and performance skill of the system during summer monsoon 2009. The study demonstrates the potential of the system for improving rainfall forecasts at five days time scale over Indian region. Districtwise performance of the ensemble rainfall forecast reveals that the technique, in general, is capable of providing reasonably good forecast skill over most states of the country, particularly over the states where the monsoon systems are more dominant.

The study provides a concise and synthesized documentation of the current level of skill of the satellite (3B42RT, 3B42V-6, KALPANA-1) products over Indian regions based on the data gathered during the summer monsoon seasons of 2006, 2007 and 2008. The inter-comparison of satellite products with the rain gauge observations suggests that the TRMM 3B42V6 product could distinctly capture characteristic features of the summer monsoon, such as north–south oriented belt of heavy rainfall along the Western Ghats with sharp gradient of rainfall between the west coast heavy rain region and the rain shadow region to the east, pockets of heavy rainfall along the location of monsoon trough, over the east central parts of the country, over north-east India, along the foothills of Himalayas and over the north Bay of Bengal. The KALPANA-1 and 3B42RT products reproduce only the broadest features of mean monsoon seasonal rainfall. The near real-time products 3B42RT and KALPANA-1 underestimate the orographic heavy rainfall along the Western Ghats of India. The precipitation estimates from TRMM 3B42V6 product, when accumulated over the whole season, could reproduce the observed pattern. However, the TRMM 3B42RT and KALPANA-1 estimates are comparatively lower than the observed rainfall over most parts of the country during the season. Inter comparison reveals that the TRMM 3B42V6 product showed better skill in estimating the daily and seasonal mean rainfall over all India and also over four homogeneous regions of India.

The study provides a concise and synthesized documentation of the current level of skill of the NCEP GFS day-1 to day-5 precipitation forecasts during Indian summer monsoon of 2008, making detailed inter-comparison with daily rainfall analysis from the use of rain gauge observations and satellite (KALPANA-1) derived Quantitative Precipitation Estimates (QPE) obtained from IMD. Model performance is evaluated for day-1 to day-5 forecasts of 24-hr accumulated precipitation in terms of several accuracy and skill measures. Forecast quality and potential value are found to depend strongly on the verification dataset, geographic region and precipitation threshold. Precipitation forecasts of the model, when accumulated over the whole season, reproduce the observed pattern. However, the model predicted rainfall is comparatively higher than the observed rainfall over most parts of the country during the season. The model showed considerable skill in predicting the daily and seasonal mean rainfall over all India and also over four broad homogeneous regions of India. The model bias for rainfall prediction changes from overestimation to underestimation at the threshold of 25 mm/day except for day-1 forecast. Model skill falls dramatically for occurrence rainfall thresholds greater than 10 mm/day. This implies that the model is much better at predicting the occurrence of rainfall than they are at predicting the magnitude and location of the peak values. Various skill score and categorical statistics for the NCEP GFS model rainfall forecast for monsoon 2008 are prepared and discussed.

During the summer monsoon (1 June to 30 September) 2007, real-time district level rainfall forecasts in short-range time scale were generated for Indian region applying multimodel ensemble technique. The pre-assigned grid point weights on the basis of correlation coefficients (CC) between the observed values and forecast values are determined for each constituent model at the resolution of 0.5° × 0.5° utilizing two seasons datasets (1 June to 30 September, 2005 and 2006), and the multimodel ensemble forecasts (day 1 and day 2 forecasts) are generated at the same resolution on a real-time basis. The ensemble forecast fields are then used to prepare forecasts for each district taking the average value of all grid points falling in a particular district. In this paper we examined the performance skill of the multimodel ensemble-based real-time district level short-range forecast of rainfall. It has clearly emerged from the results that the multimodel ensemble technique reported in this study is superior to each ensemble member. District wise performance of the ensemble rainfall forecast reveals that the technique, in general, is capable of providing reasonably good forecast skill over most districts of the country, particularly over the districts where the monsoon systems are dominant. Though the procedure shows appreciable skill to predict occurrence or non-occurrence of rainfall at the district level, it always underestimates rainfall amount, particularly in heavy rainfall events. Possible reasons of this failure may be due to model bias and poor data assimilation procedure.

The high-resolution global ensemble prediction system was implemented at IITM (Indian Institute of Tropical Meteorology) in June 2018 and handed over to India Meteorological Department (IMD) for operational running. During the Tropical Cyclone (TC) season over the North Indian Ocean, real time forecasts of ensemble tracks, intensity and strike probability of TCs are provided to IMD. This paper evaluates the skill of the model in predicting TCs during 2018, 2019 and two cases in 2020. A total of 13 cases are considered and various skill scores are calculated for the TC predictions. This study evaluates and highlights the importance of the Ensemble Prediction System for Tropical Cyclone forecasting. The key finding from this analysis is that, the higher skill of ensemble prediction system based on GEFS at longer lead time compared with the deterministic prediction. This is particularly beneficial for operational forecaster for issuing early warnings.