Estimating Daily Dew Point Temperature Based on Local and Cross-Station Meteorological Data Using CatBoost Algorithm

Abstract

Accurate estimation of dew point temperature (T dew ) plays a very important role in the fields of water resource management, agricultural engineering, climatology and energy utilization. However, there are few studies on the applicability of local T dew algorithms at regional scales. This study evaluated the performance of a new machine learning algorithm, i.e., gradient boosting on decision trees with categorical features support (CatBoost) to estimate daily T dew using limited local and cross-station meteorological data. The random forests (RF) algorithm was also assessed for comparison. Daily meteorological data from 2016 to 2019, including maximum, minimum and average temperature (T max , T min and T mean ), maximum, minimum and average relative humidity (RH max , RH min and RH mean ), maximum, minimum and average global solar radiation (Rs max , Rs min and Rs mean ) from three weather stations in Hunan of China were used to evaluate the CatBoost and RF algorithms. The results showed that both algorithms achieved satisfactory estimation accuracy at the target stations (on average RMSE = 1.020 • C, R 2 = 0.969, MAE = 0.718 • C and NRMSE = 0.087) in the absence of complete meteorological parameters (with only temperature data as input). The CatBoost algorithm (on average RMSE = 1.900 • C and R 2 = 0.835) was better than the RF algorithm (on average RMSE = 2.214 • C and R 2 = 0.828). The accuracy and stability of the CatBoost and RF algorithms were positively correlated with the number of input parameters, and the three-parameter algorithms achieved higher estimation accuracy than the two-parameter algorithms. The developed methodology is helpful to predict T dew at regional scale.

Short-Term Wind Power Prediction Method Based on Combination of Meteorological Features and CatBoost

“…CatBoost algorithm [20] can reduce the need for many super parameter tuning and the chance of over fitting, and make the model more versatile [21] . Furthermore, CatBoost can effectively avoid the impact of uneven data distribution on the model with good effect, high precision and strong generalization ability [22] .…”