GNSS-R snow depth retrieval algorithm based on PSO-LSTM

Abstract

The GNSS-R technique has been applied to retrieve snow depth, which has a high potential for application. However, in the GNSS-R classical algorithm to retrieve snow depths, the retrieval errors of high and low snow depths are large due to the influence of factors such as surface vegetation and terrain environment. In this paper, we propose a snow depth retrieval algorithm based on a particle swarm optimized long short-term memory (PSO-LSTM) neural network. The algorithm extracted three characteristic parameters (frequency, amplitude, and phase) from the SNR data as inputs, and optimized the LSTM hyperparameters by the PSO algorithm to improve the retrieval accuracy for low snow depths and snow depths close to the antenna. The snow depth retrieval results of GPS data collected from the P351 station in 2022 were evaluated in this paper. The snow depth retrieval results of the PSO-LSTM algorithm were in high agreement with the snow depth data provided by the SNOTEL network; the R-square reached 0.986, and the RMSE and MAE were 7.3 cm and 4.94 cm, respectively.
It is noteworthy that the PSO-LSTM algorithm could effectively retrieve the change in snow depth below 15cm; the value of the retrieval errors was kept within a small range. Compared with the classical algorithm, the RMSE and MAE decreased by 83.4% and 88.6%, respectively, during this period. The PSO-LSTM algorithm marginally improved the retrieval accuracy for snow depths exceeding 117 cm during this period. Compared with the classical algorithm, the RMSE and MAE decreased by 26.2% and 40.4% in this period, respectively.
In addition, the snow depth retrieval algorithm was proposed in this paper does not require antenna height and empirical formulas to realize snow depth retrieval, and at the same time, the algorithm effectively improved the retrieval accuracy for both high and low snow depths.

High-precision concentration detection of CO2 in flue gas based on BO-LSTM and variational mode decomposition

“…LSTM mainly includes many key parts, such as the forgetting gate, input gate, and output gate [30]. Its single neuron structure is shown in figure 5.…”