Major Gap Areas identified in Monsoon Mission – II

Major Gap Areas identified in Monsoon Mission - II

A. Next generation coupled seamless prediction systems

1. Develop a framework and demonstrate consistent improvement of forecasts of heavy precipitation events in GFS/NCUM at short range scale with lead time of at least 3 days. (Benchmarks for at least 4 seasons may be demonstrated)
2. Enhance the skill of sub-seasonal predictions at sub divisional scale beyond 3 weeks and achieve skill score of ~ 0.6 for seasonal forecasts, for homogenous regions over India.
3. Incorporating new methods to enhance the quality of analysis from coupled data assimilation (presently used at IITM/NCMRWF) compared to the state-of-the-art analysis products available at present with focus on land data assimilation system.
4. Global and regional Indian Ocean model development based on MOM6 including (i) high-resolution global model (~1/12°) with sea-ice module (preferably CICE5) and data assimilation, (ii) very high-resolution Indo-Pacific regional model (~1/25°) and nested (1/48°) basin-scale north Indian Ocean model with appropriate open boundary conditions, (iii) data assimilation modules in the global/regional/basin-scale model and (iv) science modules such as river discharge/freshwater flux to the ocean, open boundary conditions, mixing parameterization, biogeochemical module, etc. to augment the MOM6 based model developments.
5. Process-specific observation campaigns in the Arabian Sea for improvement of bulk flux algorithms, understanding of evolution and collapse of warm pool dynamics and understanding air-sea interaction processes in seasonal and intra-seasonal time scales.
6. Development of hybrid models with AI/ML and conventional modeling for seamless predictions in the areas of Cumulus convection, radiation and boundary layer parameterizations.

B. Development of applications based on MoES models' products

1. Improvement in the spatial and temporal scales of the forecasts reaching out to Block/Village level and for location specific forecast up to 7 days through post processing and extensive use of AI/ML techniques.
2. Generation of customized NWP products through the statistical post processing of the model outputs (deterministic and ensemble) for applications in the sectors like:
   1. QPF & ensemble based PQPF for Urban, Flash Flood & Riverine Flood
   2. Aviation weather products -a)Trend forecast during next 2-h over an airport from Model like Rainfall, winds, temperature, trend in occurrences and intensity changes of major likely weather event spells e.g. rainfall, convective events and thunderstorm etc. b)Significant weather products in map form for Asia-Icing level, wind shear product at lower level and mid-level, CAT, Turbulence at various flight levels, area of potential convection with intensity types (all products with intensity based weak, moderate and intense), level of maximum winds and jet Core. Fog Forecast products.
   3. Energy(Solar and Wind): a)Wind products at different levels of boundary layer-30, 50, 80, 100, 120 and 150m
3. Development of Extreme Weather Forecast (Thunderstorms with lightning, Hail and Squall).

In addition to the above Gap areas, MoES institutes have identified other ap areas with respect to prediction systems used at different time scales and they are listed below.

IITM
Short Range Prediction: Gap Areas

Extreme weather events, particularly intense rainfall, pose significant challenges for global and regional NWP model communities. Current models struggle to accurately predict the spatio-temporal variability of extreme rainfall events and often fail to capture the correct probability density of rainfall.

Addressing these issues is crucial for improving the reliability of weather forecasts. In this context, following gap areas were noted during the development of models under Monsoon Mission III Program by the MoES institutes.

1. In weather scale, it is noted that the ensemble forecast always remain under-dispersive or over-confident in nature while analyzing GEFS forecast.
2. It is often noted that the area/location, the time and the magnitude of extreme precipitation event is not accurately predicted by global model. Though the probabilistic forecast is found to show higher skill with longer lead (~3 days)
3. The forecast of heavy rain produced by shallow/warm rain clouds of Western Ghat mountain is generally underestimated as the mountain is not well resolved due to coarser (12 km) resolution.
4. Rapid intensification/rapid decay of cyclones remain one of the areas for further improvement. However, the cyclogenesis, track and landfall forecast by model has improved.

Extended Range Prediction: Gap Areas

1. The predictions are not skillful beyond three weeks.
2. It is difficult to predict the spatial and temporal evolution of the meteorological phenomena, especially the extreme weather events that affect smaller spatial regions, with sufficient lead time.
3. There is scarceness of reliable application-specific forecast skill evaluation/verification metrics.
4. There is a gap in communication or user awareness on the uncertainties/limitations of the forecasts, which lead to increasing user demand for accurate forecasts on smaller spatial scales on longer time scales.
5. On smaller spatial scales such as district level or sub-district level, the deterministic sub seasonal forecasts do not have value beyond a week. However, the probabilistic forecasts are skillful even up to the week4 lead. Hence the probabilistic forecasts need to be translated in a simpler way which can be understood by a layman.

Seasonal Forecasting: Gap areas

1. Predicted spatial distribution of rainfall anomalies at seasonal time scales are not skillful in present day climate models. AI/ML techniques may provide some breakthrough and needs to be explored.
2. Present CFS model coupler is hardcoded; hence developmental activities are restricted.
3. Diurnal cycle in state-of-the-art CGCMs is underestimated and MMCFS is not an exception. Diurnal rectification on to intraseasonal is a well-established research finding and hence to improve seasonal and intraseasonal predictions/simulations it is mandatory get the reasonable diurnal amplitudes/phase
4. Weakly coupled data assimilation (even at lower resolution) showed improvements in analysis; impact of high-resolution analysis on forecasts across scales needs to be established including strongly coupled data assimilation system.
5. Issuing seamless forecast and coupled NWP using MMCFS needs to be initiated and demonstrated.
6. Teleconnections (e.g IOD, Atlantic, PDO etc.) with ISMR in MMCFS are out of phase or weak, which needs to be addressed.

NCMRWF:

1. Model improvement: Proposals addressing model improvement either by post-processing or by incorporating alternative /better performing physics proceesses/parameterization in the model should be encouraged.
2. Error Tagging (Quantifying the components of the model responsible for forecast errors), Use of AI-ML, Data mining, skill evaluation of models over homogeneous regions of Indian monsoon can be considered for support.

IMD:

1. The seasonal skill of the model is poor over the Core Monsoon Zone (CMZ) and at monthly scale.
2. The skill of the model is poor for North-East Monsoon Season also.
3. The skill of prediction of Indian Ocean Dipole index is moderate.
4. Need to have a skillful LRF at State level in near future.
5. At present, HWRF modelling system is capable to track and forecast only one tropical cyclone and thereby efforts are required to have basin scale capability whereby model could also track more than one tropical cyclone simultaneously removing the requirement to run the model separately for each cyclone. The model post landfall performance is to be studied and improved thereby making the radar data ingestion necessary into the model and studies related to same are required.
6. As the model is run at very high temporal (hourly) and spatial resolution, the domains of the model do not cover the Andaman & Nicobar Islands as well as many regions of high seas. On availability of increased computational resources, the model can be extended to these important regions.
7. Also, the model needs to be further improved by assimilating newer high spatial and temporal data types such as satellite date. More products with respect to thunderstorm, lightning and extreme rainfall could be provided based on the requirement from the forecasters and other stake holders.

INCOIS:

1. The model resolution of INCOIS-GODAS is very coarse (1 degree in zonal and 1/3rd of a degree in meridional) and, thus, cannot resolve mesoscale processes in most parts of the global ocean. Also, the ocean model used in this system is Modular Ocean Model version 4 (MOM4), which is very old given that the advanced MOM6 is now available. Therefore, an upgrade in the ocean model with much higher resolution is necessary to meet the ever increasing operational requirements.
2. The Arabian Sea has a vital role in shaping climatic patterns in the Indian Ocean region's in various spatiotemporal scales. According to recent research, this body of water experiences faster warming and regional climate shifts, which could intensify extreme weather phenomenon like tropical cyclones which significantly impact the marine ecosystem and fisheries. To improve our understanding of the air-sea interaction over the Arabian Sea to the point where we can reduce systematic errors caused by misrepresentation of ocean boundary layer dynamics in models used to predict monsoon rainfall, it is proposed to conduct several field campaigns in the Arabian Sea to collect fine-scale oceanographic and atmospheric measurements, with a focus on examining surface mixed layer/interior ocean processes and marine atmospheric boundary layer processes and their representations in numerical models. In addition, it is proposed to collect one-year-long measurements using INCOIS-Flux mooring equipped with a direct covariance flux system (DCFS), meteorological sensors, the highresolution vertical profile of near-surface current, and temperature and salinity measurements in the northeastern Arabian Sea and southeastern Arabian Sea, the regions known for its pivotal role in the onset and progression of the Indian summer monsoon. This approach will facilitate an improved understanding of air-sea exchanges of heat, freshwater, and momentum at the airsea interface, ocean mixing characteristics, upper ocean response to atmospheric forcing and its impact on the upper ocean vertical structures. The DCFS facilitates direct estimates of fluxes over the wave boundary layer using the eddy covariance method. It provides an unprecedented opportunity to validate the bulk flux algorithms to estimate turbulent and momentum fluxes at the air-sea interfaces and improve the accuracy of transfer coefficients used in these schemes. These measurements will help understand the source model biases in the climate models used for seasonal and extended-range prediction and are well-versed with the overall objective of the monsoon mission. A scientific programme, Enhancing Knowledge of the Arabian Sea Marine Environment through Science and Advanced Training (EKAMSAT) is formulated as a joint research initiative between India and USA to achieve this goal. The Indian component of EKAMSAT is already projected under Monsoon Mission-III.