Rainfall–runoff modeling is a complex hydrological issue that still has room for improvement. This study developed a coupled bidirectional long short‐term memory (LSTM) with sequence‐to‐sequence (Seq2Seq) learning (BiLSTM‐Seq2seq) model to simulate multi‐step‐ahead runoff for flood events. The bidirectional LSTM with Seq2Seq learning (LSTM‐Seq2Seq) and multilayer perceptron (MLP) was set as benchmarks. The results show that: (1) root mean absolute error is reduced by approximately 19% up to 27%, and the Nash–Sutcliffe coefficient of efficiency is improved by 14% up to 34% for 6‐h‐ahead runoff prediction for BiLSTM‐Seq2Seq compared LSTM‐Seq2Seq and MLP; (2) The BiLSTM‐Seq2Seq model has good performance not only for one‐peak flood events but also for multi‐peak flood events; and (3) BiLSTM‐Seq2Seq can mitigate the time‐delay problem and time lag is shortened by 39% up to 69% in comparison to LSTM‐Seq2Seq and MLP. These results suggest that the time‐delay problem can be mitigated by BiLSTM‐Seq2Seq, which has excellent potential in time series predictions in the hydrological field.

Soil moisture (SM) has normally been estimated based on a linear relationship between SM and the surface reflectivity ($\Gamma$) from the spaceborne Global Navigation Satellite System (GNSS)-Reflectometry, while it usually relies on inputs of SM data without considering vegetation optical depth (VOD/$\tau$) effects. In this study, a new scheme is proposed for retrieving soil moisture from the Cyclone GNSS (CyGNSS) data. The variation of CyGNSS-derived $\Delta \Gamma$ is modeled as a function of both variations in SM and VOD ($\Delta \mathrm{S}\mathrm{M}$ and $\Delta \tau$). For retrieving SM, ancillary $\tau$ data can be obtained from the Soil Moisture Active Passive (SMAP) mission. In addition to this option, a model for simulating $\Delta \tau$ is suggested as an alternative. Experimental evaluation is performed for the time span from August 2019 to July 2021. Excellent agreements between the final retrievals and referenced SMAP SM products are achieved for both training (1-year period) and test (1-year duration) sets. On the whole, overall correlation coefficients (r) of 0.97 and 0.95 and root-mean-square errors (RMSEs) of 0.024 and 0.028 cm${}^3$/cm${}^3$ are obtained based on models using the SMAP and simulated $\Delta \tau$, respectively. The model without $\tau$ generates an r of 0.95 and an RMSE of 0.031 cm${}^3$/cm${}^3$. The efficiency and necessity of considering $\tau$ are thus confirmed by its enhancement based on correlation and RMSE against the one without $\tau$, and the usefulness of approximating $\Delta \tau$ by sinusoidal functions is also validated. Influences of SM statistics in terms of mean and variance on the retrieval accuracy are evaluated. This work unveils the interaction between CyGNSS data, SM, and $\tau$ and demonstrates the feasibility of integrating the $\Delta \tau$ approximation function into a bilinear regression model to obtain SM results.

Tai Lake Basin, a key freshwater resource in eastern China, has garnered attention due to widespread cyanobacterial blooms. Effective water quality management is vital for the region’s sustainable development. Investigating the seasonal variations of water quality parameters (WQPs) in Tai Lake Basin is essential for devising targeted strategies to enhance water quality. This study employs an interpretable machine learning model (XGBoost-SHAP) to identify the most important factors of water quality using daily monitoring WQP data from 2023 to 2024. Results revealed that dissolved oxygen (DO), total phosphorus (TP), permanganate index (CODMn), and ammonia nitrogen (NH₃-N) are primary determinants of water quality in the basin, while water temperature, pH, total nitrogen (TN), and turbidity showed minimal impact (SHAP value < 1). Seasonal analysis demonstrated that DO exerts a substantial influence on water quality during spring, summer, and autumn; TP and CODMn have a stable and negative impact on water quality throughout the year; NH₃-N has a relatively significant negative impact on winter water quality. Recommendations include enhancing DO levels in spring and summer, fortifying TP and NH₃-N concentrations in winter, and implementing tailored strategies in response to seasonal variations. This research offers valuable insights to guide decision-making processes aimed at enhancing water quality and safeguarding the water environment in the Tai Lake Basin.

The precipitation conce ntration degree (PCD) and precipitation concentration period (PCP) in the Liaohe River basin (LRB) from 1960 to 2020 were calculated depending on the daily precipitation data derived from meteorological stations. The mutations of the PCD and PCP were identified by sliding t-test, and spatiotemporal evolution characteristics before and after the mutation point were further analyzed. Cross wavelet transform (CWT) was used to reveal the influence of four low-frequency climate factors (Pacific Decadal Oscillation (PDO), Arctic Oscillation (AO), El Niño -Southern Oscillation (ENSO), and Sunspots (SS)) on precipitation concentration. The results were presented as follows: Mutations occurred in the PCD sequence in 1980 and the PCP sequence in 2005 in the LRB. Spatial distribution of the PCD generally increased from the southeast to the northwest and tended to flatten. Over the past 60 years, the annual PCD tended to decrease, with a variation range of 0.53 to 0.80. The PCP was relatively concentrated in early July to early August, decreasing before and increasing after the mutation. Important climatic factors driving the mutation of PCD included PDO, SS, and AO. However, the resonance between climate factors and the PCD was characterized by complexity and diversity. The PCP was mainly affected by AO and SS before the mutation. ENSO had an important influence on both PCD and PCP, but had no significant correlation with mutation occurrence.

Precipitation plays a significant role in human society and the environment, and how large‐scale climatic features influence precipitation has obtained worldwide attention. The nonhomogeneous hidden Markov model (NHMM) is a typical method to downscale large‐scale climatic elements to the regional level for many hydrologic applications. The traditional NHMM using point estimates of parameters has the risk to have no solutions for parameter estimation, but the Bayesian‐NHMM provides a Bayesian method to estimate parameters and avoids the risk. In this study, the suitability of the Bayesian‐NHMM in East Asia is evaluated. Two typical regions (i.e., the Yangtze River basin and the Zhujiang River basin in China) dominated by different climates were chosen to check model performance. Results show that: (1) the model could divide rainy‐season precipitation into several hidden states, whose variation is correspondent to the variability of monsoon and flux moisture transportation; (2) the model captures seasonality and inter‐annual variation of precipitation amount and also wet days during rainy season well in both river basins. These results suggest that the model could improve the prediction of daily precipitation in East Asia, which in turn could help many regions with similar climatic conditions worldwide to supervise floods and droughts.