Xia Wu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Xia Wu has a total of 6 co-authors. Here are the top co-authors by number of shared publications:
- Ashish Sharma
School of Civil and Environmental Engineering University of New South Wales Sydney NSW Australia
5 shared publications - Lucy Marshall
Faculty of Engineering The University of Sydney Sydney NSW Australia
4 shared publications - Qingyun Duan
College of Hydrology & Water Resources Hohai University Nanjing China
3 shared publications - Ze Jiang
School of Civil and Environmental Engineering University of New South Wales Sydney NSW Australia
2 shared publications - Zhu Liu
Hohai University
2 shared publications - Yu Lang
Institute of Natural Resource Owner’s Equity, Chinese Academy of Natural Resources Economics, Beijing 101149, China
1 shared publication
Latest Publications
- Open Access
- 18 March 2025
- Published by Copernicus GmbH
Predicting extreme precipitation events using machine learning techniques based on cumulative distribution function (CDF) analysis of meteorological factors
Extreme precipitation events have become increasingly frequent and intense in recent decades, resulting in severe flooding and substantial socio-economic losses. These events are typically associated with intense weather systems that vary across numerous meteorological factors and exhibit significant temporal and spatial variability. A comprehensive understanding of the underlying processes and the identification of key meteorological factors driving extreme precipitation are critical for enhancing the accuracy of extreme rainstorm predictions and flood warnings.This study utilized cumulative distribution function (CDF) analysis based on ERA5 hourly reanalysis data and employed the eXtreme Gradient Boosting (XGBoost) algorithm to identify the key meteorological factors contributing to 24-hour extreme precipitation across three distinct climatic zones in China. Additionally, forecasting models were developed to predict these events. The results highlighted the efficacy of this methodology and demonstrated its ability to achieve the following key advancements:Mapping data into the CDF space effectively addressed the challenges posed by the spatial heterogeneity in the value ranges of meteorological factors in regional system analyses, thereby significantly enhancing the spatial scalability of the predictive model. The integration of SHAP (SHapley Additive exPlanations) value interpretation with XGBoost successfully identified the critical meteorological factors influencing extreme precipitation events. This facilitated the construction of classification and regression models to predict both the occurrence and the return periods of these events. The application of SHAP values enhanced the interpretability of the "black-box" XGBoost model by incorporating physical insights and elucidating the interactions between different factors, thus providing valuable information for the construction and refinement of the final model. In summary, this study presents a novel and interpretable machine learning framework for analyzing and predicting extreme precipitation events based on the CDF analysis. By effectively addressing spatial heterogeneity and enhancing model interpretability, the proposed methodology offers significant advancements in the prediction of extreme rainfall and associated flood risks, contributing to improved disaster preparedness and mitigation efforts.
- 18 March 2025
- Published by Copernicus GmbH
Open AccessPredicting extreme precipitation events using machine learning techniques based on cumulative distribution function (CDF) analysis of meteorological factors
Extreme precipitation events have become increasingly frequent and intense in recent decades, resulting in severe flooding and substantial socio-economic losses. These events are typically associated with intense weather systems that vary across numerous meteorological factors and exhibit significant temporal and spatial variability. A comprehensive understanding of the underlying processes and the identification of key meteorological factors driving extreme precipitation are critical for enhancing the accuracy of extreme rainstorm predictions and flood warnings.This study utilized cumulative distribution function (CDF) analysis based on ERA5 hourly reanalysis data and employed the eXtreme Gradient Boosting (XGBoost) algorithm to identify the key meteorological factors contributing to 24-hour extreme precipitation across three distinct climatic zones in China. Additionally, forecasting models were developed to predict these events. The results highlighted the efficacy of this methodology and demonstrated its ability to achieve the following key advancements:Mapping data into the CDF space effectively addressed the challenges posed by the spatial heterogeneity in the value ranges of meteorological factors in regional system analyses, thereby significantly enhancing the spatial scalability of the predictive model. The integration of SHAP (SHapley Additive exPlanations) value interpretation with XGBoost successfully identified the critical meteorological factors influencing extreme precipitation events. This facilitated the construction of classification and regression models to predict both the occurrence and the return periods of these events. The application of SHAP values enhanced the interpretability of the "black-box" XGBoost model by incorporating physical insights and elucidating the interactions between different factors, thus providing valuable information for the construction and refinement of the final model. In summary, this study presents a novel and interpretable machine learning framework for analyzing and predicting extreme precipitation events based on the CDF analysis. By effectively addressing spatial heterogeneity and enhancing model interpretability, the proposed methodology offers significant advancements in the prediction of extreme rainfall and associated flood risks, contributing to improved disaster preparedness and mitigation efforts.
- Open Access
- 1 February 2025
- Published by IOP Publishing in Environmental Research Letters
- Vol. 20 (2) , 024049
A combined wavelet analysis-quantile mapping (WA-QM) method for bias correction: capturing the intra-annual temporal patterns in climate model precipitation simulations and projections
Despite recent improvements, global climate models (GCMs) still have biases that prevent their direct application. Quantile mapping (QM) has been widely used in bias correction (BC) due to its effectiveness in aligning the cumulative density function of climate model data with observations. However, QM has a significant limitation: it fails to consider the temporal correspondence between the model data and observations, which restricts its ability to represent intra-annual temporal patterns. To address this issue, this study integrated wavelet analysis (WA) into QM to develop a combined method termed WA-QM, which aimed to preserve QM’s efficacy in overall BC and preservation of climate change signals, while capturing the intra-annual temporal patterns. The effectiveness of WA-QM was investigated using monthly precipitation data from five Coupled Model Intercomparison Project Phase 6 models in the Pan Third Pole region, which includes Central Asia (CA), Southeast Asia, and the Tibetan Plateau. Quantile delta mapping (QDM) and scaled distribution mapping (SDM) served as the benchmark methods for assessment. The results indicated that integrating WA into QDM or SDM did not compromise the ability of QDM or SDM to correct overall biases and preserve the model’s climate change signal. Furthermore, WA could be an effective tool to overcome QM’s limitation in capturing intra-annual temporal patterns. The WA approach employed discrete wavelet transformation to decompose GCM data into various frequency bands and then adjusted their standard deviations and signs to ensure that their relative relationships were consistent with those in observations. Compared to the standalone QM approach, WA-QM showed greater accuracy in monthly statistical metrics. In the CA region, WA-SDM reduced the monthly root mean square error of the mean by 25.2%, the standard deviation by 17.6%, and the 90th percentile by 26.7% compared to SDM. The effectiveness of WA-QM was demonstrated across different data periods, spatial areas, and GCMs.
- 1 February 2025
- Published by IOP Publishing in Environmental Research Letters
- Vol. 20 (2) , 024049
Open AccessA combined wavelet analysis-quantile mapping (WA-QM) method for bias correction: capturing the intra-annual temporal patterns in climate model precipitation simulations and projections
Despite recent improvements, global climate models (GCMs) still have biases that prevent their direct application. Quantile mapping (QM) has been widely used in bias correction (BC) due to its effectiveness in aligning the cumulative density function of climate model data with observations. However, QM has a significant limitation: it fails to consider the temporal correspondence between the model data and observations, which restricts its ability to represent intra-annual temporal patterns. To address this issue, this study integrated wavelet analysis (WA) into QM to develop a combined method termed WA-QM, which aimed to preserve QM’s efficacy in overall BC and preservation of climate change signals, while capturing the intra-annual temporal patterns. The effectiveness of WA-QM was investigated using monthly precipitation data from five Coupled Model Intercomparison Project Phase 6 models in the Pan Third Pole region, which includes Central Asia (CA), Southeast Asia, and the Tibetan Plateau. Quantile delta mapping (QDM) and scaled distribution mapping (SDM) served as the benchmark methods for assessment. The results indicated that integrating WA into QDM or SDM did not compromise the ability of QDM or SDM to correct overall biases and preserve the model’s climate change signal. Furthermore, WA could be an effective tool to overcome QM’s limitation in capturing intra-annual temporal patterns. The WA approach employed discrete wavelet transformation to decompose GCM data into various frequency bands and then adjusted their standard deviations and signs to ensure that their relative relationships were consistent with those in observations. Compared to the standalone QM approach, WA-QM showed greater accuracy in monthly statistical metrics. In the CA region, WA-SDM reduced the monthly root mean square error of the mean by 25.2%, the standard deviation by 17.6%, and the 90th percentile by 26.7% compared to SDM. The effectiveness of WA-QM was demonstrated across different data periods, spatial areas, and GCMs.
- Open Access
- 20 January 2025
- Published by Copernicus GmbH
Identifying input and output data errors in the calibration of a water quality model using Bayesian error analysis with reordering (BEAR) method
The presence of errors in water quality and hydrologic variables can significantly impair the calibration of water quality models. To enhance the estimation of model parameters, it is important to accurately identify data errors during the calibration process. However, this task is challenging due to the complex interactions between model parameter uncertainty and data uncertainty. Existing methods for incorporating data uncertainty in model calibration have limitations, such as high-dimensional computation or the inability to handle stochastic errors.To address these challenges, a novel method called Bayesian Error Analysis with Reordering (BEAR) has been developed. Given that the data uncertainty arises from the data itself and is independent of the model calibration or simulation, the cumulative distribution function (CDF) of the data error can be estimated ahead and regarded as the prior information of Bayesian inference. Then the values of data error series only depend on their ranks in the CDF. BEAR method transforms the values of data error series into their ranks in the CDF. This transformation enables the effective identification of input and/or output data errors in water quality calibration.The innovation of the BEAR method can be attributed to several key aspects:1) Modification of the secant method to handle the non-linear transformation from input to output, ensuring the correspondence between the rank of input data error and the residual error of the model.2) Decomposition of model simulations to calculate the delay between each input and its corresponding output.3) Utilization of the Autoregressive model to account for the correlation of residual errors.4) Selection of an appropriate updating logic to minimize the compensation effects among multiple sources of data uncertainty.Overall, the BEAR method demonstrates flexibility and adaptability to various environmental modelling scenarios, making it a valuable tool for improving model specification under conditions of data uncertainty.
- 20 January 2025
- Published by Copernicus GmbH
Open AccessIdentifying input and output data errors in the calibration of a water quality model using Bayesian error analysis with reordering (BEAR) method
The presence of errors in water quality and hydrologic variables can significantly impair the calibration of water quality models. To enhance the estimation of model parameters, it is important to accurately identify data errors during the calibration process. However, this task is challenging due to the complex interactions between model parameter uncertainty and data uncertainty. Existing methods for incorporating data uncertainty in model calibration have limitations, such as high-dimensional computation or the inability to handle stochastic errors.To address these challenges, a novel method called Bayesian Error Analysis with Reordering (BEAR) has been developed. Given that the data uncertainty arises from the data itself and is independent of the model calibration or simulation, the cumulative distribution function (CDF) of the data error can be estimated ahead and regarded as the prior information of Bayesian inference. Then the values of data error series only depend on their ranks in the CDF. BEAR method transforms the values of data error series into their ranks in the CDF. This transformation enables the effective identification of input and/or output data errors in water quality calibration.The innovation of the BEAR method can be attributed to several key aspects:1) Modification of the secant method to handle the non-linear transformation from input to output, ensuring the correspondence between the rank of input data error and the residual error of the model.2) Decomposition of model simulations to calculate the delay between each input and its corresponding output.3) Utilization of the Autoregressive model to account for the correlation of residual errors.4) Selection of an appropriate updating logic to minimize the compensation effects among multiple sources of data uncertainty.Overall, the BEAR method demonstrates flexibility and adaptability to various environmental modelling scenarios, making it a valuable tool for improving model specification under conditions of data uncertainty.
- Open Access
- 20 January 2025
- Published by Copernicus GmbH
Precipitation bias correction: A novel method combining wavelet analysis and quantile mapping (WA-QM)
While CMIP6 has made notable improvements compared to CMIP5, there are still biases present in its simulations of different climate features due to limitations in model physics and uncertainties in input data. To address these biases, effective bias correction methods need to be employed. One commonly used method is quantile mapping, which aligns the probability density function (PDF) of climate simulations with observed data. However, this method has a limitation as it fails to maintain the temporal correspondence between model predictions and observations.To overcome this limitation, a new approach called Wavelet Analysis-Quantile Mapping (WA-QM) has been proposed. This method involves decomposing GCM simulations into different frequency bands using discrete wavelet transformation. The scaling factors of these bands are adjusted based on their correlations with observed data. Additionally, a quantile mapping procedure is applied to modify the overall PDF of the simulations.The WA-QM method was tested in monthly precipitation simulation data from five CMIP6 models covering the period 1951-2010 in the Pan Third Pole (PTP) region, which includes the Tibetan Plateau, Central Asia, and Southeast Asia. Results showed that WA-QM combines the advantages of wavelet analysis and quantile mapping. It effectively improves the representation of seasonal and monthly climatology varying through wavelet analysis, while also correcting mean and variance biases using quantile mapping. Consequently, the PDF closely resembles the observed data. The effectiveness of the WA-QM approach extends to correcting precipitation biases across different spatial areas and CMIP6 models.
- 20 January 2025
- Published by Copernicus GmbH
Open AccessPrecipitation bias correction: A novel method combining wavelet analysis and quantile mapping (WA-QM)
While CMIP6 has made notable improvements compared to CMIP5, there are still biases present in its simulations of different climate features due to limitations in model physics and uncertainties in input data. To address these biases, effective bias correction methods need to be employed. One commonly used method is quantile mapping, which aligns the probability density function (PDF) of climate simulations with observed data. However, this method has a limitation as it fails to maintain the temporal correspondence between model predictions and observations.To overcome this limitation, a new approach called Wavelet Analysis-Quantile Mapping (WA-QM) has been proposed. This method involves decomposing GCM simulations into different frequency bands using discrete wavelet transformation. The scaling factors of these bands are adjusted based on their correlations with observed data. Additionally, a quantile mapping procedure is applied to modify the overall PDF of the simulations.The WA-QM method was tested in monthly precipitation simulation data from five CMIP6 models covering the period 1951-2010 in the Pan Third Pole (PTP) region, which includes the Tibetan Plateau, Central Asia, and Southeast Asia. Results showed that WA-QM combines the advantages of wavelet analysis and quantile mapping. It effectively improves the representation of seasonal and monthly climatology varying through wavelet analysis, while also correcting mean and variance biases using quantile mapping. Consequently, the PDF closely resembles the observed data. The effectiveness of the WA-QM approach extends to correcting precipitation biases across different spatial areas and CMIP6 models.
- Open Access
Investigating the Linkage between Extreme Rainstorms and Concurrent Synoptic Features: A Case Study in Henan, Central China
Extraordinary floods are linked with heavy rainstorm systems. Among various systems, their synoptic features can be quite different. The understanding of extreme rainstorms by their causative processes may assist in flood frequency analysis and support the evaluation of any changes in flood occurrence and magnitudes. This paper aims to identify the most dominant meteorological factors for extreme rainstorms, using the ERA5 hourly reanalysis dataset in Henan, central China as a case study. Past 72 h extreme precipitation events are investigated, and six potential factors are considered in this study, including precipitable water (PW), the average temperature (Tavg) of and the temperature difference (Tdiff) between the value at 850 hPa and 500 hPa, relative humidity (RH), convective available potential energy (CAPE), and vertical wind velocity (Wind). The drivers of each event and the dominant factor at a given location are identified using the proposed metrics based on the cumulative distribution function (CDF). In Henan, central China, Wind and PW are dominant factors in summer, while CAPE and Wind are highly related factors in winter. For Zhengzhou city particularly, Wind is the key driver for summer extreme rainstorms, while CAPE plays a key role in winter extreme precipitation events. It indicates that the strong transport of water vapor in summer and atmospheric instability in winter should receive more attention from the managers and planners of water resources. On the contrary, temperature-related factors have the least contribution to the occurrence of extreme events in the study area. The analysis of dominant factors can provide insights for further flood estimations and forecasts.
Open AccessInvestigating the Linkage between Extreme Rainstorms and Concurrent Synoptic Features: A Case Study in Henan, Central China
Extraordinary floods are linked with heavy rainstorm systems. Among various systems, their synoptic features can be quite different. The understanding of extreme rainstorms by their causative processes may assist in flood frequency analysis and support the evaluation of any changes in flood occurrence and magnitudes. This paper aims to identify the most dominant meteorological factors for extreme rainstorms, using the ERA5 hourly reanalysis dataset in Henan, central China as a case study. Past 72 h extreme precipitation events are investigated, and six potential factors are considered in this study, including precipitable water (PW), the average temperature (Tavg) of and the temperature difference (Tdiff) between the value at 850 hPa and 500 hPa, relative humidity (RH), convective available potential energy (CAPE), and vertical wind velocity (Wind). The drivers of each event and the dominant factor at a given location are identified using the proposed metrics based on the cumulative distribution function (CDF). In Henan, central China, Wind and PW are dominant factors in summer, while CAPE and Wind are highly related factors in winter. For Zhengzhou city particularly, Wind is the key driver for summer extreme rainstorms, while CAPE plays a key role in winter extreme precipitation events. It indicates that the strong transport of water vapor in summer and atmospheric instability in winter should receive more attention from the managers and planners of water resources. On the contrary, temperature-related factors have the least contribution to the occurrence of extreme events in the study area. The analysis of dominant factors can provide insights for further flood estimations and forecasts.
- Open Access
- 4 March 2022
- Published by Copernicus GmbH in Hydrology and Earth System Sciences
- Vol. 26 (5) , 1203-1221
Quantifying input uncertainty in the calibration of water quality models: reordering errors via the secant method
Uncertainty in input can significantly impair parameter estimation in water quality modeling, necessitating the accurate quantification of input errors. However, decomposing the input error from the model residual error is still challenging. This study develops a new algorithm, referred to as the Bayesian Error Analysis with Reordering (BEAR), to address this problem. The basic approach requires sampling errors from a pre-estimated error distribution and then reordering them with their inferred ranks via the secant method. This approach is demonstrated in the case of total suspended solids (TSSs) simulation via a conceptual water quality model. Based on case studies using synthetic data, the BEAR method successfully improves the input error identification and parameter estimation by introducing the error rank estimation and the error position reordering. The results of a real case study demonstrate that, even with the presence of model structural error and output data error, the BEAR method can approximate the true input and bring a better model fit through an effective input modification. However, its effectiveness depends on the accuracy and selection of the input error model. The application of the BEAR method in TSS simulation can be extended to other water quality models.
- 4 March 2022
- Published by Copernicus GmbH in Hydrology and Earth System Sciences
- Vol. 26 (5) , 1203-1221
Open AccessQuantifying input uncertainty in the calibration of water quality models: reordering errors via the secant method
Uncertainty in input can significantly impair parameter estimation in water quality modeling, necessitating the accurate quantification of input errors. However, decomposing the input error from the model residual error is still challenging. This study develops a new algorithm, referred to as the Bayesian Error Analysis with Reordering (BEAR), to address this problem. The basic approach requires sampling errors from a pre-estimated error distribution and then reordering them with their inferred ranks via the secant method. This approach is demonstrated in the case of total suspended solids (TSSs) simulation via a conceptual water quality model. Based on case studies using synthetic data, the BEAR method successfully improves the input error identification and parameter estimation by introducing the error rank estimation and the error position reordering. The results of a real case study demonstrate that, even with the presence of model structural error and output data error, the BEAR method can approximate the true input and bring a better model fit through an effective input modification. However, its effectiveness depends on the accuracy and selection of the input error model. The application of the BEAR method in TSS simulation can be extended to other water quality models.
- 11 December 2021
- Published by Wiley in Hydrological Processes
- Vol. 36 (1) , e14452
Incorporating multiple observational uncertainties in water quality model calibration
The presence of error in water quality and hydrologic variables can significantly impair the calibration of water quality models. Precise and reliable identification of observational errors can have a significant impact on improving model parameter estimation. This study develops the Bayesian error analysis with reordering (BEAR) method to accommodate multiple sources of observational errors in the calibration of a water quality model. It realizes this goal by sampling the errors for input and output data from their respective error distributions and reordering them with inferred ranks via the secant method. This approach is demonstrated in the case of total suspended solids (TSS) simulated via a conceptual water quality model. Based on case studies using synthetic data, the new algorithm successfully quantifies one source of observational error when the error model of another source of observational error can be estimated accurately in advance. The results of a real case study also illustrate that considering observational errors in both model inputs and outputs, rather than in just the inputs or outputs, can improve the parameter calibration and error characterization. The improvements depend on the precision of the prior information for the error model. The application of this new algorithm in TSS simulation can be an example to understand how the BEAR algorithm works in other water quality models. The core idea of the BEAR algorithm is flexible and can be extended to the calibration of other environmental models.
- 11 December 2021
- Published by Wiley in Hydrological Processes
- Vol. 36 (1) , e14452
Incorporating multiple observational uncertainties in water quality model calibration
The presence of error in water quality and hydrologic variables can significantly impair the calibration of water quality models. Precise and reliable identification of observational errors can have a significant impact on improving model parameter estimation. This study develops the Bayesian error analysis with reordering (BEAR) method to accommodate multiple sources of observational errors in the calibration of a water quality model. It realizes this goal by sampling the errors for input and output data from their respective error distributions and reordering them with inferred ranks via the secant method. This approach is demonstrated in the case of total suspended solids (TSS) simulated via a conceptual water quality model. Based on case studies using synthetic data, the new algorithm successfully quantifies one source of observational error when the error model of another source of observational error can be estimated accurately in advance. The results of a real case study also illustrate that considering observational errors in both model inputs and outputs, rather than in just the inputs or outputs, can improve the parameter calibration and error characterization. The improvements depend on the precision of the prior information for the error model. The application of this new algorithm in TSS simulation can be an example to understand how the BEAR algorithm works in other water quality models. The core idea of the BEAR algorithm is flexible and can be extended to the calibration of other environmental models.
- Open Access
- 12 November 2020
- Published by Copernicus GmbH in EGUsphere
- Vol. 2020, 1-26
Quantifying input uncertainty in the calibration of water quality models: reshuffling errors via the secant method
Uncertainty in inputs can significantly impair parameter estimation in water quality modeling, necessitating accurate quantification of input errors. However, decomposing input error from model residual error is still challenging. This study develops a new algorithm, referred to as Bayesian error analysis with reshuffling (BEAR), to address this problem. The basic approach requires sampling errors from a pre-estimated error distribution and then reshuffling them with their inferred ranks via the secant method. This approach is demonstrated in the case of total suspended solids (TSS) simulation via a conceptual water quality model. Based on case studies using synthetic data, the BEAR method successfully isolates the input error and parameter error. The results of a real case study demonstrate that even with the presence of model structural error and output data error, the BEAR method can approximate the true input and bring a better model fit through an effective input modification. However, its effectiveness is limited by the assumption that the input uncertainty should be dominant and that the prior information of the input error model can be estimated. The application of the BEAR method in TSS simulation is effective for understanding a range of water quality conditions and the further developed algorithm can be extended to other water quality predictions.
- 12 November 2020
- Published by Copernicus GmbH in EGUsphere
- Vol. 2020, 1-26
Open AccessQuantifying input uncertainty in the calibration of water quality models: reshuffling errors via the secant method
Uncertainty in inputs can significantly impair parameter estimation in water quality modeling, necessitating accurate quantification of input errors. However, decomposing input error from model residual error is still challenging. This study develops a new algorithm, referred to as Bayesian error analysis with reshuffling (BEAR), to address this problem. The basic approach requires sampling errors from a pre-estimated error distribution and then reshuffling them with their inferred ranks via the secant method. This approach is demonstrated in the case of total suspended solids (TSS) simulation via a conceptual water quality model. Based on case studies using synthetic data, the BEAR method successfully isolates the input error and parameter error. The results of a real case study demonstrate that even with the presence of model structural error and output data error, the BEAR method can approximate the true input and bring a better model fit through an effective input modification. However, its effectiveness is limited by the assumption that the input uncertainty should be dominant and that the prior information of the input error model can be estimated. The application of the BEAR method in TSS simulation is effective for understanding a range of water quality conditions and the further developed algorithm can be extended to other water quality predictions.