This study introduces a novel deep learning model called CNN-BiLSTM for near-real-time daily wildfire spread prediction. The model integrates Convolutional Neural Network (CNN) and Bidirectional Long Short-Term Memory (BiLSTM) to capture spatial and temporal patterns in wildfire spread. The model uses the Visible Infrared Imaging Radiometer Suite (VIIRS) active fire product and various environmental variables, including topography, land cover, temperature, NDVI, wind, precipitation, soil moisture, and runoff. A comprehensive exploration of parameter configurations was conducted to optimize the model's performance. The evaluation results show that the CNN-BiLSTM model achieves an IoU of 0.58 for the validation set and 0.73 for the training set, outperforming benchmark models such as LSTM and CNN-LSTM. The model's ability to process near-real-time data and integrate spatial and temporal considerations makes it a promising tool for enhancing wildfire management efforts. The study also highlights the importance of environmental variables in wildfire spread prediction, with soil moisture showing a positive correlation with wildfire spread. The model's performance is influenced by factors such as threshold values and time steps, with optimal configurations contributing to improved prediction accuracy. The study acknowledges the model's limitations, including the need for accurate initial burn maps and the potential for false positives. Future research could focus on incorporating advanced deep learning techniques and augmenting historical wildfire data to enhance the model's predictive capabilities. The CNN-BiLSTM model demonstrates superior predictive capabilities in wildfire spread prediction, offering a significant step forward in mitigating the impact of wildfires.This study introduces a novel deep learning model called CNN-BiLSTM for near-real-time daily wildfire spread prediction. The model integrates Convolutional Neural Network (CNN) and Bidirectional Long Short-Term Memory (BiLSTM) to capture spatial and temporal patterns in wildfire spread. The model uses the Visible Infrared Imaging Radiometer Suite (VIIRS) active fire product and various environmental variables, including topography, land cover, temperature, NDVI, wind, precipitation, soil moisture, and runoff. A comprehensive exploration of parameter configurations was conducted to optimize the model's performance. The evaluation results show that the CNN-BiLSTM model achieves an IoU of 0.58 for the validation set and 0.73 for the training set, outperforming benchmark models such as LSTM and CNN-LSTM. The model's ability to process near-real-time data and integrate spatial and temporal considerations makes it a promising tool for enhancing wildfire management efforts. The study also highlights the importance of environmental variables in wildfire spread prediction, with soil moisture showing a positive correlation with wildfire spread. The model's performance is influenced by factors such as threshold values and time steps, with optimal configurations contributing to improved prediction accuracy. The study acknowledges the model's limitations, including the need for accurate initial burn maps and the potential for false positives. Future research could focus on incorporating advanced deep learning techniques and augmenting historical wildfire data to enhance the model's predictive capabilities. The CNN-BiLSTM model demonstrates superior predictive capabilities in wildfire spread prediction, offering a significant step forward in mitigating the impact of wildfires.