Search

In this study, the global ocean analysis and forecast system is evaluated against the observations in the Indian Ocean for the 2017–2018 period. A comparison of temperature and salinity analysis with seven moored buoy observations showed an excellent mutual agreement with systematically underestimating the upper ocean variability. A similar comparison of ocean analysis with the ARGO dataset average over the different oceanic regions also shows an underestimation of temperature and salinity up to 300 m depth. Further, the temperature RMSE gradually increases from surface to thermocline depth for all regions. The sea surface temperature (SST) forecast is reasonably good at all buoy locations with RMSE less than 0.5°C and a correlation of more than 0.7 up to the day-7 forecasts for the 2017–2018 period. Similarly, the sea surface salinity (SSS) forecast is also good except northern Bay of Bengal (BoB) region where the high variability leads to poor correlation and large RMSE in this shallow and freshwater region. This study helps to understand how observations fit into the model through data assimilation and also quantifies the model forecast error. Further, the study is also vital for the development/research activity related to the model and data assimilation in the Tropical Indian Ocean (TIO) region.

This study aims to evaluate the performance of five global medium-range operational NWP model rainfall forecasts, namely NCUM, UKMO, IMD GFS, NCEP GFS and ECMWF to provide an intercomparison of rainfall forecasts over India in terms of skill in predicting daily rainfall (24-hr accumulated rainfall). Verification and intercomparison of rainfall forecasts over India during monsoon 2020 (JJAS) are carried out using both (i) standard traditional verification methods (POD, FAR, RMSE, etc.) and (ii) advanced spatial verification methods (MODE, FSS). The evaluation also includes assessment of large-scale mean patterns, temporal evolution of spells during the season, dominant modes using spectral analysis, basin-scale rainfall time series and isolated heavy rainfall cases. Our analysis suggests that some of the key large-scale aspects of monsoon (seasonal mean, active/break spells, and northward propagation) are realistically represented in all the models, with slight discrepancies. In addition, the spectral analysis of rainfall is in association with observed rainfall in Day-1 forecast and deteriorates with lead times. Synoptic variance in NCUM on longer leading times is closer to observations. While the standard categorical verification over India as a whole (spatial averaged) suggests that ECMWF forecast skill is relatively high among the five models, the verification over the sub-regions shows mixed results with no clear unique higher performer among the models. In addition, basin-scale verification of rainfall forecasts for five rivers over the Indian subcontinent shows a fairly good amount of skill in terms of CC and RMSE up to Day-3 with comparable scores among the models. The advanced spatial verification metrics, like MODE and FSS, applied to the models show varying skills with different attributes. However, for FSS, forecast skill was high (low) for lower (higher) rainfall thresholds of 20 mm/day (100 mm/day). Though different models with different spatial resolutions show reasonable skill scores for larger regions, for high-impact heavy rainfall events, which are generally localised, the models have very comparable poor skill with no clear edge by a model among the five models.

Operational forecasting of tropical cyclone (TC) track and intensity in the India Meteorological Department (IMD) relies more and more on the numerical weather prediction (NWP) model guidance from national and international agencies particularly, on the medium range (24–120 h). Any improvement in TC forecasts by the NWP models enhances the operational forecaster's confidence and capability. The real-time information from the National Centre for Medium Range Weather Forecasting (NCMRWF) global NWP model (NCUM-G) is routinely used by operational forecasters at IMD as model guidance. The present study documents the improved skill of NCUM-G in forecasting the North Indian Ocean (NIO) TCs during 2015–2019, based on a collection of 1810 forecasts involving 22 TC cases. The study highlights three significant changes in the modelling system during the recent five years, namely (i) increased grid resolution from 17 to 12 km, (ii) use of hybrid 4D-Var data assimilation (DA), and (iii) increased volume of assimilated data. The study results indicate a consistent improvement in the NCUM-G model forecasts during the pre-monsoon (April–May, AM) and post-monsoon (October–December, OND) TC seasons. In addition to a 44% reduction in the initial position error, the study also reports a statistically significant decrease in the direct position error (DPE) and error in the intensity forecast, resulting in a forecast gain of 24 hrs. Comparing NWP models with IMDs official track error shows that NCUM-G and ECMWF model forecasts feature lower DPE than IMD in 2019, particularly at higher (96, 108, and 120 h) lead times.

There is a demand for reliable rainfall data-set over the South Asia region covering both land and ocean for model validation/development and various applications. For satellite rainfall estimates (SREs), the algorithm development groups also need validation information on SRE. The Tropical Rainfall Measuring Mission (TRMM) project has produced recently improved version 7 (V7) rainfall data-sets. Version 6 (V6) and V7 of 3A25, the surface rainfall products derived from TRMM precipitation radar (PR), are compared with gauge-based observations at 0.5° latitude/longitude resolution for the period of 1998–2007 over the South Asian land region. Both 3A25V7 and 3A25V6 represent the mean rainfall distribution patterns reasonably well. However, 3A25 products overestimate rainfall over the Indonesian region compared to gauge-based data. For some parts of South Asia, SREs show considerable difference in the magnitude of coefficient of variation compared to gauge-based information. At seasonal scale, a contrasting feature in bias over India during the pre-monsoon and monsoon seasons is noticed from both the versions of 3A25 data-set. In general, 3A25 rainfall data-sets are able to capture the interannual variability of rainfall over South Asia. The frequency distribution of monthly rainfall rate reveals that 3A25 products marginally underestimate rainfall below 10 mm day−1and overestimate higher rainfall rate compared to gauge-based data. Overall, 3A25V7 product is marginally better than its previous version (3A25V6) over the South Asian land region.

For Earth’s climate system, the study of the seasonal variability of sea-ice is important as the sea-ice has a significant impact on the net radiative flux, which can influence the mean seasonal behaviours of the atmosphere and ocean. In this study, the seasonal hindcast of 14 austral winter seasons is conducted to assess the skill of a coupled model in simulating the seasonal Antarctic sea-ice and its connection with the other ocean and atmospheric variables. The GloSea4 set-up of the HadGEM3 coupled model is used for the seasonal simulations at the NCMRWF. The model could reproduce the sea-ice extent over the Antarctic for the Austral winter seasons with an average correlation value of 0.98. However, there are moderate biases in the sea-ice concentration. The sea-ice thickness in the model generally shows negative bias, which is not seen to be related to the surface air temperature biases in the coupled system. The moderate positive (warm) biases in the sea surface temperature extending into the upper ocean (30 m), combined with the sea-ice drift bias pattern away from the sea-ice region are the main reasons for the underestimation of sea-ice thickness in the model. The sea surface current bias pattern shows a poleward component that brings the warm water from the warm biased locations of the exterior region into the sea-ice region and explains the presence of warmer waters in the sea-ice regions. The anti-clockwise bias in the surface wind is seen to impact the surface current, Antarctic circumpolar current (ACC), having a similar anti-clockwise current bias. Despite these moderate biases in the model, the inter-annual variability of sea-ice extent is having a reasonably good skill. The model is suitable for extended/seasonal prediction of sea-ice during Austral winter for Antarctic.