Application of Artificial Intelligence Models for Aeolian Dust Prediction at Different Temporal Scales: A Case with Limited Climatic Data

Accurately predicting ambient dust plays a crucial role in air quality management and hazard mitigation. Dust emission is complex, non-linear response to several climatic variables. This study explores the accuracy of Artificial Intelligence (AI) models: an adaptive-network-based fuzzy inference system (ANFIS) multi-layered perceptron artificial neural network (mlp-NN), over Southwestern United States (SWUS), based on observed data from IMPROVE stations. The fine (PM2.5) coarse (PM10) concentrations monthly seasonal timescales 1990–2020 are modeled using average daily maximum wind speed (W), precipitation (P), temperature (T) available North American Regional Reanalysis (NARR) dataset. model’s performance measured correlation (r), root mean square error (RMSE), percentage bias (% BIAS). ANFIS model generally performs better than mlp-NN regional dustiness SWUS region, with r = 0.77 0.83 for dust, respectively. AI models perform timescale both dust. predict timescales. Compared precipitation, more important predictor relative importance higher PM2.5 vice versa PM10. findings this demonstrate that can limited data, good potential implications research sparse regions.