Search

China has undergone rapid urbanization and internal migration in past years and its up-to-date gridded population datasets are essential for diverse applications. Existing datasets for China, however, suffer from either outdatedness or failure to incorporate the latest seventh national population census data conducted in 2020. In this study, we develop a novel population downscaling approach that leverages stacking ensemble learning and geospatial big data to produce up-to-date population grids at a 100-m resolution for China from the seventh census data at both county and town levels. The proposed approach employs random forest, XGBoost, and LightGBM as base models for stacking ensemble learning and delineates the inhabited areas from geospatial big data to enhance the gridded population estimation. Experimental results demonstrate that the proposed approach exhibits the best fit performance compared to individual base models. Meanwhile, the out-ofsample town-level test set indicates that the estimated gridded population dataset (R2=0.8936) is more accurate than existing WorldPop (R2=0.7427) and LandScan (R2=0.7165) products for China in 2020. Furthermore, with the inhabited areas enhancement, the spatial distribution of population grids is more reasonable intuitively than the two existing products. Hence, the proposed population downscaling approach provides a valuable option for producing gridded population datasets. The estimated 100-m gridded population dataset of China holds great significance for future applications and it is publicly available at https://figshare.com/s/d9dd5f9bb1a7f4fd3734 (Chen et al., 2024). [ABSTRACT FROM AUTHOR]

Accurate wind speed prediction is crucial for the safe and efficient utilization of wind resources. However, current single-value deterministic numerical weather prediction methods employed by wind farms do not adequately meet the actual needs of power grid dispatching. In this study, we propose a new hybrid forecasting method for correcting 10 m wind speed predictions made by the Weather Research and Forecasting (WRF) model. Our approach incorporates variational mode decomposition (VMD), principal component analysis (PCA), and five artificial intelligence algorithms: deep belief network (DBN), multilayer perceptron (MLP), random forest (RF), eXtreme gradient boosting (XGBoost), light gradient boosting machine (lightGBM), and the Bayesian optimization algorithm (BOA). We first predict wind speeds using the WRF model, with initial and lateral boundary conditions from the Global Forecast System (GFS). We then perform two sets of experiments with different input factors and apply BOA optimization to tune the four artificial intelligence models, ultimately building the final models. Furthermore, we compare the aforementioned five optimal artificial intelligence models suitable for five provinces in southern China in the wintertime: VMD-PCA-RF in December 2021 and VMD-PCA-lightGBM in January 2022. We find that the VMD-PCA-RF evaluation indices exhibit relative stability over nearly a year: the correlation coefficient (R) is above 0.6, forecasting accuracy (FA) is above 85 %, mean absolute error (MAE) is below 0.6 m s -1 , root mean square error (RMSE) is below 0.8 m s -1 , relative mean absolute error (rMAE) is below 60 %, and relative root mean square error (rRMSE) is below 75 %. Thus, for its promising performance and excellent year-round robustness, we recommend adopting the proposed VMD-PCA-RF method for improved wind speed prediction in models. [ABSTRACT FROM AUTHOR]

Satellite remote sensing of PM 2.5 (fine particulate matter) mass concentration has become one of the most popular atmospheric research aspects, resulting in the development of different models. Among them, the semi-empirical physical approach constructs the transformation relationship between the aerosol optical depth (AOD) and PM 2.5 based on the optical properties of particles, which has strong physical significance. Also, it performs the PM 2.5 retrieval independently of the ground stations. However, due to the complex physical relationship, the physical parameters in the semi-empirical approach are difficult to calculate accurately, resulting in relatively limited accuracy. To achieve the optimization effect, this study proposes a method of embedding machine learning into a semi-physical empirical model (RF-PMRS). Specifically, based on the theory of the physical PM 2.5 remote sensing (PMRS) approach, the complex parameter (VE f , a columnar volume-to-extinction ratio of fine particles) is simulated by the random forest (RF) model. Also, a fine-mode fraction product with higher quality is applied to make up for the insufficient coverage of satellite products. Experiments in North China (35 ∘ –45 ∘ N, 110 ∘ –120 ∘ E) show that the surface PM 2.5 concentration derived by RF-PMRS has an average annual value of 57.92 µ g m -3 vs. the ground value of 60.23 µ g m -3. Compared with the original method, RMSE decreases by 39.95 µ g m -3 , and the relative deviation is reduced by 44.87 %. Moreover, validation at two Aerosol Robotic Network (AERONET) sites presents a time series change closer to the true values, with an R of about 0.80. This study is also a preliminary attempt to combine model-driven and data-driven models, laying the foundation for further atmospheric research on optimization methods. [ABSTRACT FROM AUTHOR]

This study investigated the accuracy of the random forest (RF) model in gap filling the sensible (H) and latent heat (LE) fluxes, by using the observation data collected at a site over rice–wheat rotation croplands in Shouxian County of eastern China from 15 July 2015 to 24 April 2019. Firstly, the variable significance of the machine learning (ML) model's five input variables, including the net radiation (Rn), wind speed (WS), temperature (T), relative humidity (RH), and air pressure (P), was examined, and it was found that Rn accounted for 78 % and 76 % of the total variable significance in H and LE calculating, respectively, showing that it was the most important input variable. Secondly, the RF model's accuracy with the five-variable (Rn , WS, T , RH, P) input combination was evaluated, and the results showed that the RF model could reliably gap fill the H and LE with mean absolute errors (MAEs) of 5.88 and 20.97 W m -2 , and root mean square errors (RMSEs) of 10.67 and 29.46 W m -2 , respectively. Thirdly, four-variable input combinations were tested, and it was found that the best input combination was (Rn , WS, T , P) by removing RH from the input list, and its MAE values of H and LE were reduced by 12.65 % and 7.12 %, respectively. At last, through the Taylor diagram, H and LE gap-filling accuracies of the RF model, the support vector machine (SVM) model, the k nearest-neighbor (KNN) model, and the gradient boosting decision tree (GBDT) model were intercompared, and the statistical metrics showed that RF was the most accurate for both H and LE gap filling, while the LR and KNN model performed the worst for H and LE gap filling, respectively. [ABSTRACT FROM AUTHOR]

The alpine grassland ecosystem accounts for 53 % of the Qinghai–Tibet Plateau (QTP) area and is an important ecological protection barrier, but it is fragile and vulnerable to climate change. Therefore, continuous monitoring of grassland aboveground biomass (AGB) is necessary. Although many studies have mapped the spatial distribution of AGB for the QTP, the results vary widely due to the limited ground samples and mismatches with satellite pixel scales. This paper proposed a new algorithm using unmanned aerial vehicles (UAVs) as a bridge to estimate the grassland AGB on the QTP from 2000 to 2019. The innovations were as follows: (1) in terms of ground data acquisition, spatial-scale matching among the traditional ground samples, UAV photos, and MODIS pixels was considered. A total of 906 pairs between field-harvested AGB and UAV sub-photos and 2602 sets of MODIS pixel-scale UAV data (over 37 000 UAV photos) were collected during 2015–2019. Therefore, the ground validation samples were sufficient and scale-matched. (2) In terms of model construction, the traditional quadrat scale (0.25 m 2) was successfully upscaled to the MODIS pixel scale (62 500 m 2) based on the random forest and stepwise upscaling methods. Compared with previous studies, the scale matching of independent and dependent variables was achieved, effectively reducing the impact of spatial-scale mismatch. The results showed that the correlation between the AGB values estimated by UAV and MODIS vegetation indices was higher than that between field-measured AGB and MODIS vegetation indices at the MODIS pixel scale. The multi-year validation results showed that the constructed MODIS pixel-scale AGB estimation model had good robustness, with an average R2 of 0.83 and RMSE of 34.13 g m -2. Our dataset provides an important input parameter for a comprehensive understanding of the role of the QTP under global climate change. The dataset is available from the National Tibetan Plateau/Third Pole Environment Data Center (10.11888/Terre.tpdc.272587; H. Zhang et al., 2022). [ABSTRACT FROM AUTHOR]

With the explosive growth of atmospheric data, machine learning models have achieved great success in air pollution forecasting because of their higher computational efficiency than the traditional chemical transport models. However, in previous studies, new prediction algorithms have only been tested at stations or in a small region; a large-scale air quality forecasting model remains lacking to date. Huge dimensionality also means that redundant input data may lead to increased complexity and therefore the over-fitting of machine learning models. Feature selection is a key topic in machine learning development, but it has not yet been explored in atmosphere-related applications. In this work, a regional feature selection-based machine learning (RFSML) system was developed, which is capable of predicting air quality in the short term with high accuracy at the national scale. Ensemble-Shapley additive global importance analysis is combined with the RFSML system to extract significant regional features and eliminate redundant variables at an affordable computational expense. The significance of the regional features is also explained physically. Compared with a standard machine learning system fed with relative features, the RFSML system driven by the selected key features results in superior interpretability, less training time, and more accurate predictions. This study also provides insights into the difference in interpretability among machine learning models (i.e., random forest, gradient boosting, and multi-layer perceptron models). [ABSTRACT FROM AUTHOR]

This study investigated the accuracy of the Random Forest (RF) model in gap-filling the sensible (H) and latent heat (LE) fluxes, by using the observation data collected at a site over rice-wheat-rotation croplands in Shouxian County of eastern China from 15 July 2015 to 24 April 2019. Firstly, the variable significances of the machine learning (ML) model's five input variables, including the net radiation (Rn), winds speed (WS), temperature (T), relative humidity (RH), and air pressure (P), were examined, and it was found that Rn accounted for 78 % and 76 % of the total variable significance in H and LE calculating, respectively, showing that it was the most important input variable. Secondly, the RF model's accuracy with the five-variable (Rn, WS, T, RH, P) input combination was evaluated, and the results showed that the RF model could reliably gap-fill the H and LE with mean absolute errors (MAEs) of 5.88 Wm-2 and 20.97 Wm-2, and root mean square errors (RMSEs) of 10.67 Wm-2 and 29.46 Wm-2, respectively. Thirdly, 4-variable input combinations were tested, and it was found that the best input combination was (Rn, WS, T, P) with the MAE of H and LE reduced by 12.65 % and 7.12 %, respectively, after removing RH from the input list. At last, through the Taylor diagram, H and LE gap-filling accuracy of the RF model, the support vector machine (SVM) model, the k-nearest neighbor (KNN) model, and the gradient boosting decision tree (GBDT) model was inter-compared, and the statistical metrics showed that RF was the most accurate for both H and LE gap-filling, while the LR and KNN model performed the worst for H and LE gap-filling, respectively. [ABSTRACT FROM AUTHOR]

High-quality gridded soil moisture products are essential for many Earth system science applications, while the recent reanalysis and remote sensing soil moisture data are often available at coarse resolution and remote sensing data are only for the surface soil. Here, we present a 1 km resolution long-term dataset of soil moisture derived through machine learning trained by the in situ measurements of 1789 stations over China, named SMCI1.0 (Soil Moisture of China by in situ data, version 1.0). Random forest is used as a robust machine learning approach to predict soil moisture using ERA5-Land time series, leaf area index, land cover type, topography and soil properties as predictors. SMCI1.0 provides 10-layer soil moisture with 10 cm intervals up to 100 cm deep at daily resolution over the period 2000–2020. Using in situ soil moisture as the benchmark, two independent experiments were conducted to evaluate the estimation accuracy of SMCI1.0: year-to-year (ubRMSE ranges from 0.041 to 0.052 and R ranges from 0.883 to 0.919) and station-to-station experiments (ubRMSE ranges from 0.045 to 0.051 and R ranges from 0.866 to 0.893). SMCI1.0 generally has advantages over other gridded soil moisture products, including ERA5-Land, SMAP-L4, and SoMo.ml. However, the high errors of soil moisture are often located in the North China Monsoon Region. Overall, the highly accurate estimations of both the year-to-year and station-to-station experiments ensure the applicability of SMCI1.0 to study the spatial–temporal patterns. As SMCI1.0 is based on in situ data, it can be a useful complement to existing model-based and satellite-based soil moisture datasets for various hydrological, meteorological, and ecological analyses and models. The DOI link for the dataset is http://dx.doi.org/10.11888/Terre.tpdc.272415 (Shangguan et al., 2022). [ABSTRACT FROM AUTHOR]

Photovoltaic (PV) technology, an efficient solution for mitigating the impacts of climate change, has been increasingly used across the world to replace fossil fuel power to minimize greenhouse gas emissions. With the world's highest cumulative and fastest built PV capacity, China needs to assess the environmental and social impacts of these established PV power plants. However, a comprehensive map regarding the PV power plants' locations and extent remains scarce on the country scale. This study developed a workflow, combining machine learning and visual interpretation methods with big satellite data, to map PV power plants across China. We applied a pixel-based random forest (RF) model to classify the PV power plants from composite images in 2020 with a 30 m spatial resolution on the Google Earth Engine (GEE). The resulting classification map was further improved by a visual interpretation approach. Eventually, we established a map of PV power plants in China by 2020, covering a total area of 2917 km2. We found that most PV power plants were situated on cropland, followed by barren land and grassland, based on the derived national PV map. In addition, the installation of PV power plants has generally decreased the vegetation cover. This new dataset is expected to be conducive to policy management, environmental assessment, and further classification of PV power plants. The dataset of photovoltaic power plant distribution in China by 2020 is available to the public at 10.5281/zenodo.6849477 (Zhang et al., 2022). [ABSTRACT FROM AUTHOR]

With the explosive growth of atmospheric data, machine learning models have achieved great success in air pollution forecasting because of their higher computational efficiency than the traditional chemical transport models. However, in previous studies, new prediction algorithms have been tested only at stations or in a small region; a large-scale air quality forecasting model remains lacking to date. Huge dimensionality also means that redundant input data may lead to increased complexity and therefore the over-fitting of machine learning models. Feature selection is a key topic in machine learning development, but it has not yet been explored in atmosphere-related applications. In this work, a regional feature selection-based machine learning (RFSML) system was developed, which is capable of predicting air quality in the short-term with high accuracy at the national scale. Ensemble-Shapley additive global importance analysis is combined with the RFSML system to extract significant regional features and eliminate redundant variables at an affordable computational expense. The significance of the regional features is also explained physically. Compared with a standard machine learning system fed with relative features, the RFSML system driven by the selected key features results in superior interpretability, less training time, and more accurate predictions. This study also provides insights into the difference in interpretability among machine learning models (i.e., random forest, gradient boosting, and multi-layer perceptron models). [ABSTRACT FROM AUTHOR]

Due to rapid urbanization and intense human activities, the urban heat island (UHI) effect has become a more concerning climatic and environmental issue. A high-spatial-resolution canopy UHI monitoring method would help better understand the urban thermal environment. Taking the city of Nanjing in China as an example, we propose a method for evaluating canopy UHI intensity (CUHII) at high resolution by using remote sensing data and machine learning with a random forest (RF) model. Firstly, the observed environmental parameters, e.g., surface albedo, land use/land cover, impervious surface, and anthropogenic heat flux (AHF), around densely distributed meteorological stations were extracted from satellite images. These parameters were used as independent variables to construct an RF model for predicting air temperature. The correlation coefficient between the predicted and observed air temperature in the test set was 0.73, and the average root-mean-square error was 0.72 ∘ C. Then, the spatial distribution of CUHII was evaluated at 30 m resolution based on the output of the RF model. We found that wind speed was negatively correlated with CUHII, and wind direction was strongly correlated with the CUHII offset direction. The CUHII reduced with the distance to the city center, due to the decreasing proportion of built-up areas and reduced AHF in the same direction. The RF model framework developed for real-time monitoring and assessment of high spatial and temporal resolution (30 m and 1 h) CUHII provides scientific support for studying the changes and causes of CUHII, as well as the spatial pattern of urban thermal environments. [ABSTRACT FROM AUTHOR]

The health risks of fine particulate matter (PM 2.5) and ambient ozone (O 3) have been widely recognized in recent years. An accurate estimate of PM 2.5 and O 3 exposures is important for supporting health risk analysis and environmental policy-making. The aim of our study was to construct random forest models with high-performance and estimate daily average PM 2.5 concentration and O 3 daily maximum of 8 h average concentration (O 3 -8 hmax) of China in 2005–2017 at a spatial resolution of 1 km × 1 km. The model variables included meteorological variables, satellite data, chemical transport model output, geographic variables and socioeconomic variables. Random forest model based on 10-fold cross-validation was established, and spatial and temporal validations were performed to evaluate the model performance. According to our sample-based division method, the daily, monthly and yearly estimations of PM 2.5 from test datasets gave average model-fitting R2 values of 0.85, 0.88 and 0.90, respectively; these R2 values were 0.77, 0.77 and 0.69 for O 3 -8 hmax, respectively. The meteorological variables and their lagged values can significantly affect both PM 2.5 and O 3 -8 hmax estimations. During 2005–2017, PM 2.5 concentration exhibited an overall downward trend, while ambient O 3 concentration experienced an upward trend. Whilst the spatial patterns of PM 2.5 and O 3 -8 hmax barely changed between 2005 and 2017, the temporal trend had spatial characteristics. The dataset is accessible to the public at 10.5281/zenodo.4009308 (Ma et al., 2021a), and the shared dataset of Chinese Environmental Public Health Tracking (CEPHT, 2022) is available at https://cepht.niehs.cn:8282/developSDS3.html. [ABSTRACT FROM AUTHOR]

Deriving large-scale and high-quality precipitation products from satellite remote-sensing spectral data is always challenging in quantitative precipitation estimation (QPE), and limited studies have been conducted even using China's latest Fengyun-4A (FY-4A) geostationary satellite. Taking three rainstorm events over South China as examples, a machine-learning-based regression model was established using the random forest (RF) method to derive QPE from FY-4A observations, in conjunction with cloud parameters and physical quantities. The cross-validation results indicate that both daytime (DQPE) and nighttime (NQPE) RF algorithms performed well in estimating QPE, with the bias score, correlation coefficient and root-mean-square error of DQPE (NQPE) of 2.17 (2.42), 0.79 (0.83) and 1.77 mmh-1 (2.31 mmh-1), respectively. Overall, the algorithm has a high accuracy in estimating precipitation under the heavy-rain level or below. Nevertheless, the positive bias still implies an overestimation of precipitation by the QPE algorithm, in addition to certain misjudgements from non-precipitation pixels to precipitation events. Also, the QPE algorithm tends to underestimate the precipitation at the rainstorm or even above levels. Compared to single-sensor algorithms, the developed QPE algorithm can better capture the spatial distribution of land-surface precipitation, especially the centre of strong precipitation. Marginal difference between the data accuracy over sites in urban and rural areas indicate that the model performs well over space and has no evident dependence on landscape. In general, our proposed FY-4A QPE algorithm has advantages for quantitative estimation of summer precipitation over East Asia. [ABSTRACT FROM AUTHOR]

Due to rapid urbanization and intense human activities, the urban heat island (UHI) effect has become a more concerning climatic and environmental issue. A high spatial resolution canopy UHI monitoring method would help better understand the urban thermal environment. Taking the city of Nanjing in China as an example, we propose a method for evaluating canopy UHI intensity (CUHII) at high resolution by using remote sensing data and machine learning with a Random Forest (RF) model. Firstly, the observed environmental parameters [e.g., surface albedo, land use/land cover, impervious surface, and anthropogenic heat flux (AHF)] around densely distributed meteorological stations were extracted from satellite images. These parameters were used as independent variables to construct an RF model for predicting air temperature. The correlation coefficient between the predicted and observed air temperature in the test set was 0.73, and the average root-mean-square error was 0.72 #176;C. Then, the spatial distribution of CUHII was evaluated at 30-m resolution based on the output of the RF model. We found that wind speed was negatively correlated with CUHII, and wind direction was strongly correlated with the CUHII offset direction. The CUHII reduced with the distance to the city center, due to the de-creasing proportion of built-up areas and reduced AHF in the same direction. The RF model framework developed for real-time monitoring and assessment of high-resolution CUHII provides scientific support for studying the changes and causes of CUHII, as well as the spatial pattern of urban thermal environments. [ABSTRACT FROM AUTHOR]

High loadings of nitrate (NO3-) in the aerosol over China significantly exacerbate the air quality and pose a great threat to ecosystem safety through dry–wet deposition. Unfortunately, limited ground-level observation data make it challenging to fully reflect the spatial pattern of NO3- levels across China. Until now, long-term monthly particulate NO3- datasets at a high resolution were still missing, which restricted the assessment of human health and ecosystem safety. Therefore, a unique monthly NO3- dataset at 0.25 ∘ resolution over China during 2005–2015 was developed by assimilating surface observations, satellite products, meteorological data, land use types and other covariates using an ensemble model combining random forest (RF), gradient-boosting decision tree (GBDT), and extreme gradient-boosting (XGBoost) methods. The new developed product featured an excellent cross-validation R2 value (0.78) and relatively lower root-mean-square error (RMSE: 1.19 µgNm-3) and mean absolute error (MAE: 0.81 µgNm-3). Besides, the dataset also exhibited relatively robust performance at the spatial and temporal scales. Moreover, the dataset displayed good agreement with (R2=0.85 , RMSE=0.74 µgNm-3 , and MAE=0.55 µgNm-3) some unlearned data collected from previous studies. The spatiotemporal variations in the developed product were also shown. The estimated NO3- concentration showed the highest value in the North China Plain (NCP) (3.55±1.25 µgNm-3); followed by the Yangtze River Delta (YRD) (2.56±1.12 µgNm-3), Pearl River Delta (PRD) (1.68±0.81 µgNm-3), and Sichuan Basin (1.53±0.63 µgNm-3), and the lowest one in the Tibetan Plateau (0.42±0.25 µgNm-3). The higher ambient NO3- concentrations in the NCP, YRD, and PRD were closely linked to the dense anthropogenic emissions. Apart from the intensive human activities, poor terrain condition might be a key factor for the serious NO3- pollution in the Sichuan Basin. The lowest ambient NO3- concentration in the Tibetan Plateau was contributed by the scarce anthropogenic emission and favourable meteorological factors (e.g. high wind speed). In addition, the ambient NO3- concentration showed a marked increasing tendency of 0.10 µgNm-3yr-1 during 2005–2014 (p<0.05), while it decreased sharply from 2014 to 2015 at a rate of -0.40 µgNm-3yr-1 (p<0.05). The ambient NO3- levels in Beijing–Tianjin–Hebei (BTH), YRD, and PRD displayed gradual increases at a rate of 0.20, 0.11, and 0.05 µgNm-3yr-1 (p<0.05) during 2005–2013, respectively. The gradual increases in NO3- concentrations in these regions from 2005 to 2013 were due to the fact that the emission reduction measures during this period focused on the reduction of SO2 emission rather than NOx emission and the rapid increase in energy consumption. Afterwards, the government further strengthened these emission reduction measures and thus caused the dramatic decreases in NO3- concentrations in these regions from 2013 to 2015 (p<0.05). The long-term NO3- dataset over China could greatly deepen the knowledge about the impacts of emission reduction measures on air quality improvement. The monthly particulate NO3- levels over China during 2005–2015 are open access at 10.5281/zenodo.3988307 (Li et al., 2020c). [ABSTRACT FROM AUTHOR]

The health risks of fine particulate matter (PM2.5) and ambient ozone (O3) have been widely recognized in recent years. An accurate estimate of PM2.5 and O3 exposures is important for supporting health risk analysis and environmental policy-making. The aim of our study was to construct random forest models with high-performance, and estimate daily average PM2.5 concentration and O3 daily maximum 8h average concentration (O3-8hmax) of China in 2005-2017 at a spatial resolution of 1km×1km. The model variables included meteorological variables, satellite data, chemical transport model output, geographic variables and socioeconomic variables. Random forest model based on ten-fold cross validation was established, and spatial and temporal validations were performed to evaluate the model performance. According to our sample-based division method, the daily, monthly and yearly simulations of PM2.5 gave average model fitting R² values of 0.85, 0.88 and 0.90, respectively; these R² values were 0.77, 0.77, and 0.69 for O3-8hmax, respectively. The meteorological variables and their lagged values can significantly affect both PM2.5 and O3-8hmax simulations. During 2005-2017, PM2.5 exhibited an overall downward trend, while ambient O3 experienced an upward trend. Whilst the spatial patterns of PM2.5 and O3-8hmax barely changed between 2005 and 2017, the temporal trend had spatial characteristic. The dataset is accessible to the public at https://doi.org/10.5281/zenodo.4009308 (Ma et al., 2021), and the shared data set of Chinese Environmental Public Health Tracking: CEPHT (https://cepht.niehs.cn:8282/developSDS3.html). [ABSTRACT FROM AUTHOR]

Land cover (LC) determines the energy exchange, water and carbon cycle between Earth's spheres. Accurate LC information is a fundamental parameter for the environment and climate studies. Considering that the LC in China has been altered dramatically with the economic development in the past few decades, sequential and fine-scale LC monitoring is in urgent need. However, currently, fine-resolution annual LC dataset produced by the observational images is generally unavailable for China due to the lack of sufficient training samples and computational capabilities. To deal with this issue, we produced the first Landsat-derived annual China land cover dataset (CLCD) on the Google Earth Engine (GEE) platform, which contains 30 m annual LC and its dynamics in China from 1990 to 2019. We first collected the training samples by combining stable samples extracted from China's land-use/cover datasets (CLUDs) and visually interpreted samples from satellite time-series data, Google Earth and Google Maps. Using 335 709 Landsat images on the GEE, several temporal metrics were constructed and fed to the random forest classifier to obtain classification results. We then proposed a post-processing method incorporating spatial–temporal filtering and logical reasoning to further improve the spatial–temporal consistency of CLCD. Finally, the overall accuracy of CLCD reached 79.31 % based on 5463 visually interpreted samples. A further assessment based on 5131 third-party test samples showed that the overall accuracy of CLCD outperforms that of MCD12Q1, ESACCI_LC, FROM_GLC and GlobeLand30. Besides, we intercompared the CLCD with several Landsat-derived thematic products, which exhibited good consistencies with the Global Forest Change, the Global Surface Water, and three impervious surface products. Based on the CLCD, the trends and patterns of China's LC changes during 1985 and 2019 were revealed, such as expansion of impervious surface (+ 148.71 %) and water (+ 18.39 %), decrease in cropland (- 4.85 %) and grassland (- 3.29 %), and increase in forest (+ 4.34 %). In general, CLCD reflected the rapid urbanization and a series of ecological projects (e.g. Gain for Green) in China and revealed the anthropogenic implications on LC under the condition of climate change, signifying its potential application in the global change research. The CLCD dataset introduced in this article is freely available at 10.5281/zenodo.4417810 (Yang and Huang, 2021). [ABSTRACT FROM AUTHOR]

Deriving large-scale and high-quality precipitation products from satellite remote sensing spectral data is always challenging in quantitative precipitation estimation (QPE), and limited studies have been conducted even using the China's latest Fengyun-4A (FY-4A) geostationary satellite. Taking three rainstorm events over South China as examples, a Random Forest (RF) model framework for FY-4A QPE during daytime/nighttime is established by using FY-4A multi-band spectral information, cloud parameters, high-density precipitation observations, and physical quantities from reanalysis data. During daytime (nighttime), the probability of detection of the RF model for precipitation is 0.99 (0.99), while the correlation coefficient and root-mean-square error between the retrieved and observed precipitation are 0.77 (0.82) and 1.84 (2.32) mm/h, respectively, indicating that the RF model of FY-4A QPE has high precipitation retrieval accuracy. In particular, the RF model exhibits good spatiotemporal predictive ability for precipitation intensities within the range of 0.5–10 mm/h. For the retrieved accumulated precipitation, the precipitation intensity exhibits a greater impact on the predictive ability of the QPE algorithm than the precipitation duration. Due to the higher density of automatic stations in urban areas, the accuracy of FY-4A QPE over such areas is higher compared with rural areas. Both the accumulated precipitation and the distribution density of automatic stations are more important factors for the predictive ability of the RF model of FY-4A QPE. In general, our proposed FY-4A QPE algorithm has advantages for near-real-time monitoring of summer precipitation over East Asia. [ABSTRACT FROM AUTHOR]

Eastern China is one of the most economically developed and densely populated areas in the world. Due to its special geographical location and climate, eastern China is affected by different weather systems, such as monsoons, shear lines, typhoons, and extratropical cyclones. In the near future, the rainfall rate becomes difficult to predict precisely due to these systems. Traditional physics-based methods such as numerical weather prediction (NWP) tend to perform poorly on nowcasting problems due to the spin-up issue. Moreover, various meteorological stations are distributed in this region, generating a large amount of observation data every day, which have great potential for application to data-driven methods. Thus, it is important to train a data-driven model from scratch that is suitable for the specific weather situation of eastern China. However, due to the high degrees of freedom and nonlinearity of machine learning algorithms, it is difficult to add physical constraints. Therefore, with the intention of using various kinds of data as a proxy for physical constraints, we collected three kinds of data (radar, satellite, and precipitation data) in the flood season from 2017 to 2018 in this area and preprocessed them into tensors (256×256) that cover eastern China with a domain of 12.8×12.8∘. The developed multisource data model (MSDM) combines the optical flow, random forest, and convolutional neural network (CNN) algorithms. It treats the precipitation nowcasting task as an image-to-image problem, which takes radar and satellite data with an interval of 30 min as inputs and predicts radar echo intensity with a lead time of 30 min. To reduce the smoothing caused by convolutions, we use the optical flow algorithm to predict satellite data in the following 120 min. The predicted radar echoes from the MSDM together with satellite data from the optical flow algorithm are recursively implemented in the MSDM to achieve a 120 min lead time. The MSDM predictions are comparable to those of other baseline models with a high temporal resolution of 6 min. To solve blurry image problems, we applied a modified structural similarity (SSIM) index as a loss function. Furthermore, we use the random forest algorithm with predicted radar and satellite data to estimate the rainfall rate, and the results outperform those of the traditional, nonlinear radar reflectivity factor and rainfall rate (Z – R) relationships that use logarithmic functions. The experiments confirm that machine learning with multisource data provides more reasonable predictions and reveals a better nonlinear relationship between radar echo and precipitation rate. Apart from developing complicated machine learning algorithms, exploiting the potential of multisource data will yield more improvements. [ABSTRACT FROM AUTHOR]

Land cover (LC) determines the energy exchange, water and carbon cycle between Earth's spheres. Accurate LC information is a fundamental parameter for the environment and climate studies. Considering that the LC in China has been altered dramatically with the economic development in the past few decades, sequential and fine-scale LC monitoring is in urgent need. However, currently, fine-resolution annual LC dataset produced by the observational images is generally unavailable for China due to the lack of sufficient training samples and computational capabilities. To deal with this issue, we produced the first Landsat-derived annual China Land Cover Dataset (CLCD) on the Google Earth Engine (GEE) platform, which contains 30 m annual LC and its dynamics of China from 1990 to 2019. We first collected the training samples by combining stable samples extracted from China's Land-Use/Cover Datasets (CLUD), and visually-interpreted samples from satellite time-series data, Google Earth and Google Map. Using 335,709 Landsat images on the GEE, several temporal metrics were constructed and fed to the random forest classifier to obtain classification results. We then proposed a post-processing method incorporating spatial-temporal filtering and logical reasoning to further improve the spatial-temporal consistency of CLCD. Finally, the overall accuracy of CLCD reached 79.31% based on 5,463 visually-interpreted samples. A further assessment based on 5,131 third-party test samples showed that the overall accuracy of CLCD outperforms that of MCD12Q1, ESACCI_LC, FROM_GLC, and GlobaLand30. Besides, we intercompared the CLCD with several Lands at derived thematic products, which exhibited good consistencies with the Global Forest Change, the Global Surface Water, and three impervious surface products. Based on the CLCD, the trends and patterns of China's LC changes during 1985 and 2019 were revealed, such as expansion of impervious surface (+148.71%) and water (+18.39%), decrease of cropland (- 4.85%) and grassland (-3.29%), increase of forest (+4.34%). In general, CLCD reflected the rapid urbanization and a series of ecological projects (e.g., Gain for Green) in China and revealed the anthropogenic implications on LC under the condition of climate change, signifying its potential application in the global change research. The CLCD dataset introduced in this article is freely available at http://doi.org/10.5281/zenodo.4417810 (Yang and Huang, 2021). [ABSTRACT FROM AUTHOR]

The satellite remote-sensing aerosol optical depth (AOD) and meteorological elements were employed to invert PM2.5 in order to control air pollution more effectively. This paper proposes a restricted gradient-descent linear hybrid machine learning model (RGD–LHMLM) by integrating a random forest (RF), a gradient boosting regression tree (GBRT), and a deep neural network (DNN) to estimate the concentration of PM2.5 in China in 2019. The research data included Himawari-8 AOD with high spatiotemporal resolution, ERA-5 meteorological data, and geographic information. The results showed that, in the hybrid model developed by linear fitting, the DNN accounted for the largest proportion, whereas the weight coefficient was 0.62. The R² values of RF, GBRT, and DNN were reported 0.79, 0.81, and 0.8, respectively. Preferably, the generalization ability of the mixed model was better than that of each sub-model, and R² reached 0.84, whereas RMSE and MAE were reported 12.92 µg/m³ and 8.01 µg/m³, respectively. For the RGD-LHMLM, R² was above 0.7 in more than 70% of the sites, whereas RMSE and MAE were below 20 µg/m³ and 15 µg/m³, respectively, in more than 70% of the sites due to the correlation coefficient having seasonal difference between the meteorological factor and PM2.5. Furthermore, the hybrid model performed best in winter (mean R² was 0.84) and worst in summer (mean R² was 0.71). The spatiotemporal distribution characteristics of PM2.5 in China were then estimated and analyzed. According to the results, there was severe pollution in winter with an average concentration of PM2.5 being reported 62.10 µg/m³. However, there was slight pollution in summer with an average concentration of PM2.5 being reported 47.39 µg/m³. The findings also indicate that North China and East China are more polluted than other areas and that their average annual concentration of PM2.5 was reported 82.68 µg/m³. Moreover, there was relatively low pollution in Inner Mongolia, Qinghai, and Tibet, for their average PM2.5 concentrations were reported below 40 µg/m³. [ABSTRACT FROM AUTHOR]

Accurate and timely maps of urban underlying land properties at the national scale are of significance in improving habitat environment and achieving sustainable development goals. Urban impervious surface (UIS) and urban green space (UGS) are two core components for characterizing urban underlying environments. However, the UIS and UGS are often mosaicked in the urban landscape with complex structures and composites. The "hard classification" or binary single type cannot be used effectively to delineate spatially explicit urban land surface property. Although six mainstream datasets on global or national urban land use and land cover products with a 30 m spatial resolution have been developed, they only provide the binary pattern or dynamic of a single urban land type, which cannot effectively delineate the quantitative components or structure of intra-urban land cover. Here we propose a new mapping strategy to acquire the multitemporal and fractional information of the essential urban land cover types at a national scale through synergizing the advantage of both big data processing and human interpretation with the aid of geoknowledge. Firstly, the vector polygons of urban boundaries in 2000, 2005, 2010, 2015 and 2018 were extracted from China's Land Use/cover Dataset (CLUD) derived from Landsat images. Secondly, the national settlement and vegetation percentages were retrieved using a sub-pixel decomposition method through a random forest algorithm using the Google Earth Engine (GEE) platform. Finally, the products of China's UIS and UGS fractions (CLUD-Urban) at a 30 m resolution were developed in 2000, 2005, 2010, 2015 and 2018. We also compared our products with six existing mainstream datasets in terms of quality and accuracy. The assessment results showed that the CLUD-Urban product has higher accuracies in urban-boundary and urban-expansion detection than other products and in addition that the accurate UIS and UGS fractions were developed in each period. The overall accuracy of urban boundaries in 2000–2018 are over 92.65 %; and the correlation coefficient (R) and root mean square errors (RMSEs) of UIS and UGS fractions are 0.91 and 0.10 (UIS) and 0.89 and 0.11 (UGS), respectively. Our result indicates that 71 % of pixels of urban land were mosaicked by the UIS and UGS within cities in 2018; a single UIS classification may highly increase the mapping uncertainty. The high spatial heterogeneity of urban underlying covers was exhibited with average fractions of 68.21 % for UIS and 22.30 % for UGS in 2018 at a national scale. The UIS and UGS increased unprecedentedly with annual rates of 1605.56 and 627.78 km2yr-1 in 2000–2018, driven by fast urbanization. The CLUD-Urban mapping can fill the knowledge gap in understanding impacts of the UIS and UGS patterns on ecosystem services and habitat environments and is valuable for detecting the hotspots of waterlogging and improving urban greening for planning and management practices. The datasets can be downloaded from 10.5281/zenodo.4034161 (Kuang et al., 2020a). [ABSTRACT FROM AUTHOR]

East China is one of the most economically developed and most densely populated areas in the world. Due to its special geographical location and climate, East China is affected by different weather systems like monsoon, shear line, typhoon and extratropical cyclone, in the imminent future the rainfall rate affected by which is difficult to precisely predict. Traditional physics-based methods like Numerical Weather Prediction (NWP) tend to perform poorly for the nowcasting problem due to its spinup issue. Meanwhile, various meteorological stations are distributed here, generating a large amount of observation data every day, which has a great potential to be applied to data-driven methods. Thus, it is important to train a data-driven model from scratch that suitable to the specific weather situation of East China. We collect three kinds of data (radar, satellite, precipitation) in flood season from 2017 to 2018 of this area and preprocess them into ndarray (256 × 256) that cover East China with a domain of 12.8 × 12.8°.The Multi-Source Data Model (MSDM) which we developed combines the Optical flow, Random forest and Convolutional Neural Network (CNN). It treats the precipitation nowcasting task as an image-to-image problem, which takes radar and satellite data with a interval of 30 minutes as inputs and predicts radar echo intensity at a lead time of 30 minutes. To reduce the smoothing caused by convolution, we use Optical flow to predict satellite data in the following 120 minutes. The predicted radar echo from MSDM together with satellite data from Optical flow are recursively implemented in MSDM to achieve 120 minutes lead time. The predictions from MSDM are comparable to those of other baseline models with a high temporal resolution of 6 minutes. To solve the blurry image problems, we applied a modified SSIM as a loss function. Furthermore, we use Random forest with predicted radar and satellite data to estimate the rainfall rate, the results outperform those of the traditional Z-R relationship. The experiments confirm that machine learning with multi-source data provides more reasonable predictions and reveals a better non-linear relationship between radar echo and precipitation rate. Besides the algorithms will be developed, exploiting the potential of multi-source data will bring more improvements. [ABSTRACT FROM AUTHOR]

High loadings of nitrate (NO3-) in the aerosol over China significantly exacerbates the air quality and poses a great threaten on ecosystem safety through dry/wet deposition. Unfortunately, limited ground-level observation data makes it challenging to fully reflect the spatial pattern of NO3- level across China. Up to date, the long-term monthly NO3- datasets at a high resolution were still missing, which restricted the assessment of human health and ecosystem safety. Therefore, a unique monthly NO3- dataset at 0.25° resolution over China during 2005-2015 was developed by assimilating surface observation, satellite product, meteorological data, land use types and other covariates using an ensemble model combining random forest (RF), gradient boosting decision tree (GBDT), and extreme gradient boosting (XGBoost). The new developed product featured excellent cross-validation R2 value (0.78) and relatively lower root-mean-square error (RMSE: 1.19 μg/m3) and mean absolute error (MAE: 0.81 μg/m3). Besides, the dataset also exhibited relatively robust performance at the spatial and temporal scale. Moreover, the dataset displayed good agreement with (R2 = 0.85, RMSE = 0.74 μg/m3, and MAE = 0.55 μg/m3) some unlearning data collected from previous studies. The spatiotemporal variations of the developed product were also shown. The estimated NO3- concentration showed the highest value in North China Plain (NCP) (3.55 ± 1.25 μg/m3), followed by Yangtze River Delta (YRD (2.56 ± 1.12 g/m3)), Pearl River Delta (PRD (1.68 ± 0.81 μg/m3)), Sichuan Basin (1.53 ± 0.63 μg/m3), and the lowest one in Tibetan Plateau (0.42 ± 0.25 μg/m3). The higher ambient NO3- concentrations in NCP, YRD, and PRD were closely linked to the dense anthropogenic emissions. Apart from the intensive human activities, poor terrain condition might be a key factor for the serious NO3- pollution in Sichuan Basin. The lowest ambient NO3- concentration in Tibetan Plateau was contributed by the scarce anthropogenic emission and favorable meteorological factors (e.g., high wind speed). In addition, the ambient NO3- concentration showed marked increasing tendency of 0.10 μg/m3/year during 2005-2014 (p < 0.05), while it decreased sharply from 2014 to 2015 at a speed of -0.40 μg/m3/year (p < 0.05). The ambient NO3- levels in Beijing-Tianjin-Hebei (BTH), YRD, and PRD displayed gradual increases at the speed of 0.13, 0.08, and 0.03 μg/m3/year (p < 0.05) during 2005-2014, respectively. The gradual increases of NO3- concentrations in these regions from 2005 to 2014 were due to that the emission reduction measures during this period focused on the reduction of SO2 emission rather than NOx emission and the rapid increase of energy consumption. Afterwards, the government further strengthened these emission reduction measures, and thus caused the dramatic decreases of NO3- concentrations in these regions from 2014 to 2015 (p < 0.05). The long-term NO3- dataset over China could greatly deepen the knowledge about the impacts of emission reduction measures on air quality improvement. The monthly particulate NO3- levels over China during 2005-2015 are open access in https://doi.org/10.5281/zenodo.3988307 (Li et al., 2020c). [ABSTRACT FROM AUTHOR]