the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 07 Aug 2024
IITM High-Resolution Global Forecast Model Version 1: An attempt to resolve monsoon prediction deadlock
Abstract. The prediction of Indian monsoon rainfall variability affecting a country with a population of billions remained one of the major challenges of the numerical weather prediction community. While in recent years, there has been a significant improvement in predicting the synoptic scale transients associated with the monsoon circulation, the intricacies of rainfall variability remained a challenge. Here, an attempt is made to develop a global model using a dynamic core of a cubic octahedral grid that provides a higher resolution of 6.5 km over the global tropics. This high-resolution model has been developed to resolve the monsoon convection. Reforecasts with the IITM High-resolution Global Forecast Model (HGFM) have been run daily from June through September 2022. The HGFM model has a wave number truncation of 1534 in the cubic octahedral grid. The monsoon events have been predicted with a ten-day lead time. The HGFM model is compared to the operational GFS T1534. While the HGFM provides skills comparable to the GFS, it shows better skills for higher precipitation thresholds. This model is currently being run in experimental mode and will be made operational.
- Preprint
(2208 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on gmd-2024-89', Anonymous Referee #1, 02 Nov 2024
Review of “IITM High-Resolution Global Forecast Model Version 1: An attempt
to resolve monsoon prediction deadlock” by Krishna et al. submitted to GMD
Overview:
The authors introduced a new model that can significantly improve the prediction of monsoon, which is key to local agriculture, economy, and disaster preparedness. The prediction ability of this new model is encouraging. I think the manuscript is clearly written although some loose ends need to be addressed. Specifically, I found that the overall quality of the figures varies a lot. High-quality figures are important to convey key results. I think a major revision is needed to address this. Then it could be published at GMD. Please see my detailed comments below.
- Figure 1c, the legend needs to be fixed (solid lines vs. dash lines)
- Also figure 1c, it is inconvenient to compare the ratio of dynamics time to total time. Perhaps only using one single y-axis?
- Why only considering 200 hPa kinetic energy? How about other vertical levels?
- Figure 3: It would be better to add the model names to each panel. I am confused by lines 188-189. To me, HGFM and GFS look more similar to each other while ERA5 look quite different, especially over the gulf of Mexico and northwestern Pacific. I suggesting plotting the difference for better comparisons among the three outputs.
- Figure 4 and line 194 to 203: there are some discussions about the biases over the tropical ocean. I am wondering if there are any specific reasons why both models overestimate the precipitation over the tropical eastern Pacific? Is it due to the shallow convection scheme?
- Figure 5&6: is it cm/day or mm/day?
- Figure 7: improve the quality of this figure (mixed font sizes, panel sizes, etc.).
- Why convective precipitation is reduced from GFS to HGFM? Due to Tuning?
- What is the point of figure 13?
Citation: https://doi.org/10.5194/gmd-2024-89-RC1 -
RC2: 'Comment on gmd-2024-89', Anonymous Referee #2, 30 Nov 2024
The manuscript proposes an improvement to the GFS numerical prediction model by employing a dynamic core based on a cubic octahedral grid, enabling an increase in model resolution to 6.5 km over the tropics. The model is then applied to resolve monsoon convection, producing daily runs from June to September 2022. The results are compared to those obtained using the operational GFS T1534. The authors find that their modeling approach performs significantly better, particularly for longer lead times and heavier rain events.
This is an interesting paper, as the suggested method appears to improve weather forecasts in regions with intense precipitation. Its content aligns well with the scope of Geoscientific Model Development (GMD) journal. However, I believe several improvements are necessary before the manuscript can be considered for publication.
General Comments
-
Reference to Previous Work
Based on a web search, it seems that enhancing the resolution of numerical weather prediction models is not a novel concept and has already been applied (e.g., https://doi.org/10.1002/qj.958). However, such references are missing from the manuscript, and I suggest that the authors update the reference list accordingly. -
Clarification of Innovative Contribution
Given the above observation, the innovative aspect of this work appears to be the application of a dynamic core using a cubic octahedral grid specifically within the GFS model, rather than the general application of this method to numerical weather prediction models. If this is the case, it should be clearly stated in the text. -
Validation Period
The authors mention (lines 86–87) that the motivation for improving numerical weather forecasts in India under heavy precipitation conditions stems from difficulties in forecasting extreme rainfall over Kerala during August 2018 and August 2019. Therefore, it is unclear why they did not validate their approach for these events, opting instead for the June–September 2022 period. Additionally, limiting the validation to a single year raises concerns. It would be valuable to see how the model performs for another year or in different periods. -
Linguistic Improvements
The manuscript contains several syntax errors, which should be corrected, preferably by a native English speaker. Below are a few examples:a. Line 108: "This A new grid" → "This new grid"
b. Line 146: "...Tco model is in general..." → "...the Tco model is in general..."
c. Line 148: "...of the model run has been discussed..." → "...of the model run is discussed..."
d. Line 149: "...few case studies have also been discussed" → "...few case studies are also discussed"
e. Line 185: "...methodology suggested..." → "...the methodology suggested..."
f. Line 226: "...is vertical..." → "...is the vertical..."
Specific Comments
-
Lines 103–104
The authors state:"This paper is the first attempt, to the best of our knowledge, towards building a model close to a convection-permitting global weather model in India, with an emphasis on Indian monsoon rainfall variability."
It would be helpful if the authors clarified whether the improved performance of their model is expected to hold only for this region or if it could be effective in other geographical areas under similar conditions. If so, they should specify what these conditions might be. -
Lines 154–160
The authors compare forecasts with daily precipitation data from satellite products, gridded data from the India Meteorological Department (IMD), and the ERA5 database. However, given that previous studies highlight discrepancies between gridded data and ground measurements, it would be valuable to include a comparison with observations from ground-based meteorological stations. These are likely available in abundance for this region. -
Figure 2
The authors should discuss the discrepancies between the curves observed for wave numbers greater than 10−4 m−1 and whether these differences pose challenges to their analysis. -
Figure 3
The differences between dCAPE values from ERA5 and HGFM should be quantified, as visual inspection alone may not suffice to identify discrepancies. Additionally, the units of dCAPE should be provided next to the color code for clarity.
Citation: https://doi.org/10.5194/gmd-2024-89-RC2 -
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 283 | 80 | 16 | 379 | 7 | 7 |
- HTML: 283
- PDF: 80
- XML: 16
- Total: 379
- BibTeX: 7
- EndNote: 7
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1