Abstract:The regional Climate-Weather Research and Forecasting model (CWRF) is used to downscale the National Climate Center’s operational short-term climate prediction based on the Beijing Climate Center Climate System Model (BCC_CSM1.1m). The prediction skill is assessed for daily precipitation and surface air temperature (2 m) from March to August during 1991—2010. As compared to BCC_CSM1.1m, CWRF predicts more realistic seasonal mean precipitation and temperature spatial distributions and resolves more detailed features over mountainous regions with large terrain variations. It also predicts interannual variations in precipitation East and Central China more accurately. Overall, CWRF outperforms BCC_CSM1.1m in predicting precipitation of different intensities, especially extreme events. Benefiting from higher spatial resolution and advanced physical process simulation at lower levels, CWRF downscaling can improve the seasonal prediction of summer precipitation in China.

Abstract:Based on the outputs of the high-resolution regional climate model (RegCM4) driven by the results of five global climate system models, the simulation performance of RegCM4 for precipitation extremes in Southwest China is systematically evaluated. Furthermore, the future precipitation extremes in Southwest China are also evaluated. The results indicate that the RegCM4 models can reasonably reproduce the climate means for the period from 1986 to 2005, but there are greater biases over Southwest China, especially over central Sichuan and the Sichuan basin. Thus, bias correction is implemented on the RegCM4 outputs, and it can significantly reduce the bias of precipitation extremes over Southwest China. For future changes in precipitation extremes, the total wet day precipitation (Prcptot), very heavy precipitation days (R10mm), maximum day precipitation (Rx1day), and extreme precipitation (R95p) are projected to increase significantly across Southwest China over the 21st century under the representative concentration pathways (RCP) 4.5 and 8.5 scenarios. Compared to the current climate (1986—2005), all the precipitation extreme indices exhibit an increasing trend in the future. Both Rx1day and R95p are projected to increase in most areas of Southwest China under the scenarios of RCP4.5 and RCP8.5 in the 21st century, and are reported to respectively increase by approximately 16.0% and 12.6% at the end of the 21st century under the RCP4.5 scenario. The future changes in Prcptot and R10mm show great regionality, with both showing a decreasing trend in southern Yunnan and the Sichuan Basin, while the other regions show an increasing trend. Additionally, these changes in Prcptot and R10mm under the RCP8.5 scenario are considerably higher than those under the RCP4.5 scenario.

Abstract:High-resolution model simulation is considered one of the important methods for studying climate change over the Tibetan Plateau (TP), which is characterized with scarce observations. The High Resolution Model Intercomparison Project (HighResMIP) has been added to the Sixth International Coupled Model Comparison Program (CMIP6), but its simulation performance has not been systematically evaluated over the TP. In this study, we evaluate the ability of CMIP6 HighResMIP models to simulate historical climate over the TP and perform an ensemble projection of the TP climate trend in the near future. The results show that, when compared to lower-resolution simulations, higher-resolution simulations of almost all (two-thirds) models reduce the area-mean bias of annual mean precipitation (surface air temperature). A combined assessment of the Taylor diagram involving indices shows that higher-resolution simulations of about one-third of the models outperform their lower-resolution simulations for both annual mean surface air temperature and precipitation, while higher-resolution simulations of the rest of the models are close to or inferior to their lower-resolution simulations. Multi-model ensemble results outperform individual model results, and their higher-resolution simulation generally outperforms the lower-resolution simulation. Under the SSP5-8.5 scenario, an ensemble of higher-resolution models projects significant warming over the TP during 2021—2040 compared to 1995—2014, with relatively weak warming in the southeastern part. Projected precipitation shows an increasing-decreasing-increasing pattern from north to south. The annual mean surface air temperature will increase by (0.81± 0.91) ℃ and precipitation will increase by (0.05± 0.25) mm/d on average over the TP. These findings are useful for understanding the impact of improved model resolution on climate simulation performance over the TP and the evolution of the TP's climate in a warming future world.

Abstract:This paper proposes a combination model based on CEEMDAN-SE-ARIMA that aims to address the shortcomings of traditional time series models that cannot effectively predict modal aliased data. The proposed model combines the advantages of the complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), the high short-term prediction accuracy of an auto-regressive integrated moving average model (ARIMA), and the fast efficiency of sample entropy (SE) reconstruction. The model is empirically analyzed for summer precipitation in Northeast China from 2016 to 2020. First, based on the fully adaptive ensemble empirical mode decomposition method, the precipitation time series is decomposed into multiple eigenmode components, and the component sequence is reconstructed according to the calculation results of the entropy of different component samples. Then, for each reconstruction component, an autoregressive moving average forecast model is constructed. Finally, the predicted value of each component is superimposed to obtain the predicted value of the combined model. Additionally, the ARIMA single model and other combined modelsare constructed to be compared with the CEEMDAN-SE-ARIMA combined model. The results show that the CEEMDAN-SE-ARIMA combined accounts for the time series’ modal aliasing characteristics, effectively improves the forecasting ability of the summer precipitation time series model in Northeast China, and has good forecast application value. Compared with the single model and other combined models, the forecast results are improved. MASE decreases by 0.02—0.91 mm, RMSE decreases by 0.80—130.49 mm, MAE decreases by 2.52—129.84 mm, and MAPE decreases by 1.08—35.53 mm. The CEEMDAN-SE-ARIMA model has a better prediction effect in the northwest region, where the precipitation variability is small, and the prediction of the extreme value distribution center in the southeast region is more accurate.

Abstract:Based on the three-source fusion grid precipitation analysis data from CMPAS and using the dichotomy classical verification method, a comprehensive map of precipitation forecast score, and the Method for Object-Based Diagnostic Evaluation (MODE), we compare and analyzed the precipitation forecast performance of the fine-gridded SCMOC, ECMWF global, and CMA-Meso models in Qinling and its surrounding areas in 2021, and observe the following: 1) The ECMWF model can well describe the spatial distribution characteristics of daily average precipitation, daily precipitation standard deviation, and daily precipitation frequency under the influence of terrain. However, the precipitation frequency of more than 0.1 mm is much higher than the observation, and the torrential rain frequency is lower than the observation. SCMOC and CMA-Meso have better forecasts of precipitation of different grades. The deficiency of SCMOC is that its ability to describe the fine spatial distribution characteristics of precipitation is relatively weak. 2) The occurrence time of the daily peak of precipitation frequency greater than 0.1mm in the ECMWF model is about 3 hours earlier than the observation, and CMA-Meso and SCMOC are more consistent with the observation. 3) The TS scores of the three products with 24-hour precipitation greater than or equal to 0.1 mm are basically the same, but the characteristics of the precipitation forecast are significantly different. SCMOC has a high success rate, a low hit rate, more missed hits, and fewer false alarms than ECMWF and CMA-Meso models which are the opposite of SCMOC. SCMOC's TS score, success rate, and hit rate for 24 h, 3 h heavy rain, and above are better than the other two products. 4) The verification results of the MODE method show that SCMOC has the highest similarity between the forecast and observation of large-area precipitation objects, and its forecast ability is better than ECMWF and CMA-Meso. However, there is a high risk of missing a hit for scattered, small-area torrential rain objects. The east-west distance deviation of SCMOC and ECMWF is greater than that of the north-south direction, and the proportion of the west position is higher than that of the east position.

Abstract:This study is based on the daily 24-to 168-hour ensemble precipitation forecast datasets derived from the European Centre for Medium-Range Weather Forecasts and extracted from the TIGGE (The Interactive Grand Global Ensemble) dataset. The ensemble forecast comprises 51 ensemble members. The study applies the left-censored non-homogeneous logistic regression method (CNLR) and the standardized model post-processing method (SAMOS) to calibrate the precipitation forecasts in Southeast China. The results show that the CNLR method can effectively reduce the mean absolute error (MAE) and continuous ranked probability score (CRPS) of the raw ensemble forecast, and improve the forecasting skills of quantitative and probabilistic precipitation forecasts. Using the SAMOS method to preprocess the data and considering the impact of topography and other factors, the forecast error caused by the terrain influence can be further corrected on the basis of the CNLR method, thereby obtaining a more accurate probabilistic forecast of precipitation.

Abstract:During June 14th—17th, 2021, an extreme rainstorm process with a daily maximum of 106.6 mm occurred over the Tarim Basin. This study analyzes the synergistic mechanism between the westerly wind and Indian monsoon on this rainstorm process using precipitation data from an automatic meteorological station in the basin, the GRAPES-GFS analysis field, and ERA-5 (0.25° × 0.25° , hourly) reanalysis data, as well as the WRF-v4.2. 2 numerical simulation model and the HYSPLIT-v4.0 water vapor backward trajectory model. Under the circulation background of “high geopotential height on the east side and low on the west side” at 100 hPa and “one low trough between two high-pressure ridges” at 500 hPa, water vapor is transported from Black Sea, Caspian Sea, Salt Lake, northern Indian Ocean, Central Asia and northern Xinjiang, corresponding to westerly, southerly, and easterly water vapor tracks, each transported along 850—300 hPa, 500—400 hPa, and below 650 hPa. The Indian monsoon circulation plays a key role in transporting water vapor from the Indian Ocean to the Tarim Basin, elucidating the physical process of the southward movement of the Indian monsoon carrying water vapor into the basin. The disappearance of the “south wind window” water vapor transport is an important manifestation of the synergistic influence of westerly winds and the Indian monsoon.

Abstract:Background error covariance plays an essential role in data assimilation systems, particularly in variational assimilation systems. The National Meteorological Center (NMC) method has widely been used to generate forecast error samples for estimating background error covariance. Currently, most variational-based rapid update and cycling (RUC) data assimilation and forecasting systems use a fixed background error covariance at each analysis moment to reduce computational costs. However, with the increasing frequency of assimilation in the RUC data assimilation and forecasting systems, a fixed background error covariance may not be suitable for all analysis moments. To adopt diurnal background error covariance in the RUC data assimilation and forecasting system more reasonably, the diurnal background error covariance characteristics in summer and winter are analyzed by the NMC method based on the RMAPS-ST system, and assimilation and forecast experiments are conducted. The results show that the background error covariances in summer and winter exhibit obvious diurnal characteristics. The standard deviation of forecast error samples and the eigenvalues of each control variable (U, V, T, and RH_s) are higher at night than during the day, indicating that the forecast errors of the model system are more significant at night than during the day. Meanwhile, the standard deviation of forecast error samples and the eigenvalues of each control variable are higher in summer than in winter, suggesting that the model forecast errors of the system are greater in summer than in winter. The horizontal length scale is generally larger in summer than in winter, which may be because the spatial integrity of the RMAPS-ST system forecast error is more consistent in summer and the horizontal correlation is higher, leading to a larger length scale. The 3-day cycling experiments initially indicate that the use of diurnal background error covariances can improve the assimilation and forecast of the U, V, T, and Q fields of RMAPS-ST system, thereby enhancing the performance of precipitation forecasts.

Abstract:Based on the WRF-simulated results, surface observation data, radar data, and the ECMWF ERA-5 reanalysis data, this paper analyzes the characteristics of local frontogenesis and its influence on the development of convective systems during the heavy rainstorm occurred on October 1—8, 2010, in Hainan Island. The results show that the environmental field played a major role in the frontogenesis process of this heavy rain event. The non-adiabatic heating term and the horizontal motion term contribute the most to the process of local frontogenesis, and the largest positive value area of the two overlaps many times in the heavy precipitation area, indicating that diabatic heating and horizontal deformation divergence are responsible for the strong frontogenesis in this area. In addition, the comparison and analysis of simulated results and observations show that the lower condensation height leads to strong latent heat release in the lower convective layer during the period of the strongest precipitation, and that the internal warming of the lower air mass in the convective zone creates a strong frontogenesis effect. The strong frontogenesis at the lower level accelerates the updraft, strengthens the development of deep convection, and intensifies the torrential rain. Compared with the diabatic heating term and the horizontal movement term, the contribution of the vertical motion tilting term related to vertical movement is small, but it increases at night. The analysis shows that the difference in the horizontal distribution of vertical velocity at night plays an important role in the enhancement of deep convection in areas with heavy rainfall.

Abstract:Based on satellite-derived snow depth (SD) over the Tibetan Plateau (TP), daily rainfall data from 261 meteorological stations in South China (SC), and the ERA-5 reanalysis dataset during 1979—2018, the relationships between TP winter snow depth and precipitation during the first rainy season (FRS) in SC are investigated in this study. The results show that: 1) The connections between SD over the western TP and precipitation during the FRS in SC are the most robust, and TP SD mainly affects frontal precipitation during the FSR, whereas it shows less impact on summer monsoonal rainfall. 2) The onset date of the FRS in above-normal TP snow years is about 20 days earlier than that in below-normal snow years, leading to more rainy days, a longer FRS, and more rainfall during the FRS. However, rainfall intensity during the FRS shows little difference between different types of TP snow years. 3) The TP is colder in an above-normal snow year, and the cooling effect stimulates abnormal anticyclonic circulation over the TP. However, tripole anomaly patterns of the 500 hPa geopotential height occur in the East Asian coastal region. The circulation configurations facilitate cold air invading SC in middle-high latitude regions, making SC colder. The enhanced northwest Pacific subtropical high intensifies the low-level southerly flow and water vapor supplement. The front swings northward-southward in northern SC during March and April. The FRS is established once the dry, cold, northerly flow and the warm, wet, southerly flow invade the SC in early April. In below-normal TP snow years, both the cold northerly and warm southerly flow are weak and inactive, and the front over northern SC is interrupted in early April. When the northerly and southerly flow invades the SC in mid-to-late April, the front is reverted and the FRS in the SC occurs later.

Abstract:This study establishes a homogenized daily maximum and minimum temperature series for Baoding in Hebei Province based on daily observations collected by the National Meteorological Information Centre from 1919 to 2019. Quality control is conducted to remove error data resulting from manual observation or recording, instrument malfunctions and digital inputs. Then, missing observations are interpolated using the standardized series method with homogenized daily temperature data in Tianjin over a century-long scale. Significant breakpoints caused by changes in interpolation, station relocation, and instrument manufacturers are detected using a penalized maximal t-test (PMT) with annual and monthly reference series constructed by two means, and adjusted by Quantile Matching (QM) daily reference series from Berkeley Earth-daily data. The characteristics of inter-annual, decadal, and trend changes are consistent with those from Berkeley Earth-monthly, CRUTS 4.03, and GHCNV3. The warming change induced by the rapid urban development in the Baoding region is well reflected when simultaneously compared with the whole Beijing-Tianjin-Hebei region. Furthermore, the warming trends of the annual and autumn lowest minimum temperature (TNn) are 0.340 ℃/(10 a) and 0.404 ℃/(10 a) (95% significance level), respectively, and the corresponding diurnal temperature range (DTR) are -0.118 ℃/(10 a) and -0.215 ℃/(10 a) (95% significance level). The warming change in annual and seasonal temperature extremes in Baoding have also increased significantly since 1912.

Abstract:In this paper, an ensemble QPE is generated in order to improve the effect of estimation of the Quantitative Precipitation Estimation (QPE) in Jiangsu Meteorological Observatory, by combining the spatial structure of error uncertainty represented by the error covariance matrix and the temporal structure represented by the time lag correlation coefficient. The ensemble is generated using data from the Jiangsu Meteorological Observatory from May to August 2019 and May to August 2020, with its deterministic and probabilistic components being verified using the corresponding observed precipitation. It is found that the number of components has little effect on the ensemble QPE and that one can control component numbers between 16 and 50. The deterministic verification results show that ensemble QPE aggravates the underestimation of precipitation in some areas, but reduces the absolute error and root mean square error of precipitation in general. Ensemble mean can improve the accuracy and reduce the rate of false positives, and it will also increase the number of true positives. The ensemble has a good Brier score for precipitation of various magnitudes, with larger precipitation magnitudes having a better estimation effect. Also, the dispersion of the ensemble is small, and after the components of the set are sorted, the frequency of observations falling at both ends increases, which also reflects the small dispersion. In addition, the observed values are more frequent than the maximum values of ensemble components, which indicates that the ensemble QPE tends to underestimate precipitation. With the increase of the probability threshold, the hit rate (POD) and false alarm rate (POFD) of ensemble precipitation estimation gradually increase, but the degree of POD increase is much greater than that of POFD, resulting in a broken ROC curve. POD and POFD with different probability thresholds show that ensemble QPE has high estimation skills for all levels of precipitation, with light rain and moderate rain having the best resolution. It is more concerning that the occurrence probability of light rain and heavy rain estimated by the ensemble is less than the actual frequency, whereas the occurrence probability of moderate rain and heavy rain estimated by the ensemble is very close to the actual frequency of precipitation, which has high reliability; however, the ensemble QPE still tends to underestimate the occurrence probability of precipitation.