The goal of forest fire modeling is to understand and forecast the behavior of forest fires through numerical simulation. To simulate fire dangers and fire spread behavior, statistical techniques are applied to historical fire incidents. A number of fire-based data tables are used to create the Canadian fire weather index (FWI). FWI is a straightforward method that effectively identifies vegetation's fire vulnerability. Simple empirical equations are used to calculate the Australian fire danger index (FDI). Similar to FWI is FDI. Models for forest fires that combine stochastic and mathematical techniques accurately mimic the spread of the fire and forecast its severity. Physical models use heat transport through convection and radiation as well as the heat balance of sources (fire) and sinks (forest areas close to the fire source). MATLAB or Delphi programs are used to solve the differential equations relating to heat transfer. The world crown fire modeling experiment's predicted fire spread behavior closely matches the predictions made by mathematical approaches.  In real-time forestry, forest landscape, wind speed, ambient temperature and rain fall level are obtained from satellite imagery and weather forecasting and machine learning is used for predicting the vulnerability to forest fire and the rate of spreading of the fire.