| 06 Jul 2026
A Numerical Weather Prediction Model-Based Approach to Assess Fire Weather Conditions over the Northwest Himalayan Forests
Abstract. In recent years, forest fire activities have increased in frequency and intensity over the Indian Himalayan region. Year 2024 witnessed numerous fires spread across the Himalayan states of India, causing devastating economic and environmental impacts. The sparse observational network across the Himalayas significantly limits the availability of real-time data, thereby constraining the timely dissemination of wildfire early warnings. This study elaborates on utilising an NWP model, such as the Weather Research and Forecasting Model, for the simulation of fire weather variables for the 2024 summer fire season across the northwestern Himalayan states. A very high-resolution WRF model is configured with the NCEP-FNL reanalysis dataset as the initial and boundary conditions, and simulations are carried out to reconstruct high-resolution fire weather conditions during the 2024 fire weather season across 24 identified fire clusters. The analysis suggests that the two major fire weather indicators (1) Vapour Pressure Deficit and (2) Fire weather indices from the Canadian Fire Danger Rating System blended with NWP model simulations could be a potential tool in identifying fire weather conditions for data sparse complex terrains and subsequently issuing fire weather alerts on a daily basis where the current Fire Danger rating system operates at 10-day intervals.
This study uses the WRF model with a two‑nest configuration (finest grid at 1 km) to simulate meteorological conditions during the May 2024 fire season over the northwestern Himalayas, and compares the performance of two planetary boundary layer (PBL) schemes, YSU and MYNN. The simulations are evaluated against automatic weather station observations and ERA5‑Land reanalysis data, aiming to provide a reference for fire‑weather modelling and operational applications in complex terrain. Fire‑weather simulation over mountainous regions is an important direction in current forest‑fire research, and the topic has practical operational value. However, the current manuscript focuses primarily on comparing PBL schemes, with insufficiently refined scientific questions, limited depth in physical‑mechanism analysis, and inadequate support for fire‑risk simulation.
Major comments
Minor comments
(1) Please further supplement the WRF model configuration details – for example, whether this set of physics parameterisations has been evaluated in this region (including references, operational requirements, or computational constraints). Also, it is necessary to specify the height information corresponding to the 47 vertical levels.
(2) Since ERA5‑Land itself has uncertainties in complex mountainous areas, please discuss these uncertainties and explain why both automatic station data and ERA5‑Land were used for validation.
(3) Please add a map showing the spatial distribution of observation stations with their elevations, and also include a vegetation map. The large elevation differences in complex mountainous areas mean that station representativeness has a significant impact on evaluation results; the discussion should incorporate elevation and vegetation when analysing simulation errors. Additionally, most stations are concentrated in the UK – why is that?
(4) This study does not actually validate fire‑risk levels against actual fire points or burned areas. Therefore, it can only demonstrate the capability of different PBL schemes for simulating meteorological variables, and cannot directly prove their better suitability for fire‑risk simulation. I suggest adding case studies of actual fire events to strengthen the conclusions.
(5) What are the reasons for the large differences between simulations and observations shown in Figures S19 and S20?
(6) There are several formatting issues, such as misaligned paragraphs (e.g., lines 464, 360, 520).
(7) Figures 1 and 2 should be merged.