Precipitation amount (PA), frequency (PF), and intensity (PI) over China are characterized and quantified using a high-resolution merged satellite-gauge precipitation product for 6 years (January 2008 through December 2013). The precipitation product synthesizes both state-of-the-art multisatellite precipitation algorithms and the latest, densest gauge observations to provide high-quality precipitation information at a very fine temporal and spatial resolution (0.1°/hourly) that encompasses all of China. The geographical and seasonal variations in precipitation are systematically documented over seven subregions, each corresponding to a unique climate regime. PA, PF, and PI have large seasonal and geographical variations across China. It is found that 1) although heavy precipitation events (>10 mm/h) represent only 0.8% of total precipitation occurrence over China, they contribute 12.1% of the total precipitation volume. Light precipitation events (<;1 mm/h) dominate the precipitation occurrence (74.3%) and contribute 23.1% of the total precipitation volume; 2) over the high-altitude Tibetan Plateau (TP), the land-locked Xinjiang (XJ) province, and northwestern China (NW), light precipitation events (<;1 mm/h) occur very frequently (74.7%, 82.1%, and 64.1% of all precipitation events) and contribute 29.8%, 35.5%, and 27.4% of the total precipitation volume. This initial continental-scale study provides new insights on precipitation characteristics that can benefit meteorological and hydrological modeling and applications, especially in areas with sparse rain-gauge coverage.
The constant conflict between the rapidly developing socioeconomic and ecological environment within the Guangdong–Hong Kong–Macao Greater Bay Area necessitates the exploration of ecosystem vulnerability patterns and driving mechanisms. A comprehensive social–economic–ecological framework is proposed to assess the ecosystem vulnerability pattern of the Greater Bay Area, specifically spanning from 1990 to 2020. Employing geographic detectors and weighting methods, the study quantifies spatiotemporal variation and the underlying mechanisms driving vulnerability in the study area. The results demonstrate an obvious trend in ecosystem vulnerability index (ESVI) across the Greater Bay Area, with an initial decline followed by gradual increase during the1990–2020. A substantial majority of the region (approximately 63.85% of the total area) experienced a decline in ESVI from 1990 to 2010. Moreover, the spatial distribution of this decline exhibited a prevailing east-to-west pattern, indicating an overall southward shift over time. Furthermore, a decline was primarily concentrated in the central region and the rapidly expanding urban areas situated on both sides of the Pearl River estuary. Encouragingly, a notable amplification in ESVI was observed between 2010 and 2020, which is attributed to the development, utilization, and protection of land, forests, water bodies, and other pertinent factors associated with urban expansion. The impact of climate change on ESVI changes exhibits a growing magnitude over time, while human activities persist as the predominant driver of ESVI changes. The natural factors exerted a substantial impact on ESVI changes primarily in the upper reaches of the Pearl River, which included topographic relief, precipitation, water network density, biological abundance, and related aspects. Conversely, the pronounced influence of human activities on ESVI changes predominantly manifests within the urban agglomeration of the Pearl River Delta. Key contributors to such a manifestation encompass land change types, intensity of human activities, population density, and related variables. Changes in land use have the potential to induce heightened ecological vulnerability changes. The amelioration of ecological protection and land use practices can be mitigated and reduced by employing ESVI. Moreover, the framework introduced in this study holds the potential to extend vulnerability assessments to other regions with similar ecosystem types. It is expected that the findings derived from this framework could contribute to the formulation of policy recommendations pertaining to ecosystem protection and management.
[1] The spatial error structure of surface precipitation derived from successive versions of the TRMM Multisatellite Precipitation Analysis (TMPA) algorithms are systematically studied through comparison with the Climate Prediction Center Unified Gauge daily precipitation Analysis (CPCUGA) over the Continental United States (CONUS) for 3 years from June 2008 to May 2011. The TMPA products include the version-6(V6) and version-7(V7) real-time products 3B42RT (3B42RTV6 and 3B42RTV7) and research products 3B42 (3B42V6 and 3B42V7). The evaluation shows that 3B42V7 improves upon 3B42V6 over the CONUS regarding 3 year mean daily precipitation: the correlation coefficient (CC) increases from 0.85 in 3B42V6 to 0.92 in 3B42V7; the relative bias (RB) decreases from −22.95% in 3B42V6 to −2.37% in 3B42V7; and the root mean square error (RMSE) decreases from 0.80 in 3B42V6 to 0.48 mm in 3B42V7. Distinct improvement is notable in the mountainous West especially along the coastal northwest mountainous areas, whereas 3B42V6 (also 3B42RTV6 and 3B42RTV7) largely underestimates: the CC increases from 0.86 in 3B42V6 to 0.89 in 3B42V7, and the RB decreases from −44.17% in 3B42V6 to −25.88% in 3B42V7. Over the CONUS, 3B42RTV7 gained a little improvement over 3B42RTV6 as RB varies from −4.06% in 3B42RTV6 to 0.22% in 3B42RTV7. But there is more overestimation with the RB increasing from 8.18% to 14.92% (0.16–3.22%) over the central US (eastern).
Abstract The purpose of this study is to demonstrate the capability of an experimental, weather‐adaptive, high‐resolution, deterministic Warn‐on‐Forecast (WoF) analysis and forecast system (WoF3DVAR‐AFS) for predicting high‐impact severe weather events that occurred during the Hazardous Weather Testbed 2019 Spring Forecast Experiments. WoF3DVAR‐AFS uses a three‐dimensional variational (3DVAR) method as its core data assimilation system and the Advanced Research Version of the Weather Research and Forecasting (WRF‐ARW) model as its forward model. Surface measurements provided in meteorological aviation reports and the Oklahoma Mesonet, Doppler radar data, and spaceborne total lightning observations provided by the Geostationary Lightning Mapper are assimilated at 15‐min frequency over a target domain determined by the “Day 1” Convective Outlook product from the Storm Prediction Center. The chief goal of this system is to complement probabilistic forecasts generated by ensemble analysis and forecast systems, such as the experimental Warn‐on‐Forecast System (WoFS) with a higher‐resolution deterministic member to aid forecasters' decision‐making. We performed both qualitative and quantitative evaluations on 0–6 hr forecasts launched hourly from 1900 to 0300 UTC the next day for each of the 12 cases. Aggregated subjective forecast evaluation metrics from each individual case, as well as detailed comparison against available verification datasets, suggest that the forecasts are generally skillful in terms of composite reflectivity fields, quantitative precipitation forecasts, and the strength and location of rotation tracks and damaging winds. This study presents initial efforts to assess the performance of WoF3DVAR‐AFS and provides possible directions for further improvements, including the development of a weather‐adaptive, dual‐resolution analysis and forecast system hybrid with an ensemble system, such as the experimental Warn‐on‐Forecast system.
This study develops an objective deep-learning-based model for tropical cyclone (TC) intensity estimation. The model’s basic structure is a convolutional neural network (CNN), which is a widely used technology in computer vision tasks. In order to optimize the model’s structure and to improve the feature extraction ability, both residual learning and attention mechanisms are embedded into the model. Five cloud products, including cloud optical thickness, cloud top temperature, cloud top height, cloud effective radius, and cloud type, which are level-2 products from the geostationary satellite Himawari-8, are used as the model training inputs. We sampled the cloud products under the 13 rotational angles of each TC to augment the training dataset. For the independent test data, the model shows improvement, with a relatively low RMSE of 4.06 m/s and a mean absolute error (MAE) of 3.23 m/s, which are comparable to the results seen in previous studies. Various cloud organization patterns, storm whirling patterns, and TC structures from the feature maps are presented to interpret the model training process. An analysis of the overestimated bias and underestimated bias shows that the model’s performance is highly affected by the initial cloud products. Moreover, several controlled experiments using other deep learning architectures demonstrate that our designed model is conducive to estimating TC intensity, thus providing insight into the forecasting of other TC metrics.
Precipitation is the critical components in the hydrological cycle over the Earth. Accurate and reliable high-resolution precipitation nowcasting with up to two hours ahead is a strong demand for many sectors that make decisions based on weather forecast. Due to the complexity of atmosphere, precipitation prediction is a long standing scientific challenge with direct impact on social and economic departments throughout the world. Numerical Weather Prediction(NWP) is the most common approach to provide people with weather information in near future based on the simulation of atmospheric dynamics based on physical laws, which usually comes across the "spin-up" problem and is hard to make full use of tremendous amount of Earth observation data that increase quickly as more and more observation instruments are being deployed around the world. The machine learning (ML) networks (DNN) witnessed remarkable progress in recent years due to increased amounts of available data, better model architectures and ease of implementation on powerful specialized hardware such as GPU and TPUs (Sφnderby 2020; Jouppi et al., 2017). Recently the DNN is now able to effectively process large time series spatial and temporal input data. Up on these developments of ML and hardware, we introduce China Artificial Intelligence Nowcasting system(CAINS) based on a deep learning model for precipitation nowcasting over China. As shown in Figure 1, at least 8 temporally continuous radar observed images are needed to drive CAINS to produce 60 (six-hour) precipitation images. Figure 2 shows the 6-hour precipitation nowcasting results with initial time at 20210513(Beijing time, BJT) 08:00. Figure 3 show the performance of nowcasting results by CAINS and Optical Flow(OF) algorithm in terms of Probability Of Detection (POD), Critical Success Index (CSI), and Equitable Threat Score(ETS). It is noted that CAINS outperform OF with pronouncedly better skills.
The continued presence of systematic errors in operational forecasts of return flow over the Gulf of Mexico has motivated an investigation into this problem. The theme of the work is use of a low-order mixed-layer model that is faithful to the phenomenon in the context of dynamic data assimilation. Data assimilation experiments in the identical-twin mode determine the best place to make observations that minimize the forecast error through adjustment of model controls. The emphasized controls are those associated with the fluxes of heat and moisture from sea to atmosphere. Results indicate that the best observations are at that time and place when the outflowing continental air passes over the warmest sea surface temperatures. In the case studied, this warmest zone is directly over the Loop Current. Observations at times long after the modified air leaves these warmest waters lead to relatively poor control adjustments and little improvement in the forecast. If input to data assimilation is restricted to observations of a single model variable over short intervals of time (the order of several hours), results are relatively poor. Yet, a significant improvement is forthcoming if one of the observations is replaced by an observation from another model variable. This result is understood through arguments based on forecast sensitivity to model control. The paper ends with discussion of steps to be taken that hold promise for correcting systematic error in return-flow forecasts.
Abstract Building on the results from the observing system simulation experiments in Part I, this study investigates the impact of assimilating Geostationary Operational Environmental Satellite‐16 (GOES‐16) derived atmospheric motion vector (AMV) data on the convective scale numerical weather prediction (NWP) by using the National Severe Storms Laboratory (NSSL) three‐dimensional variational (3DVAR) data assimilation (DA) system. The benefit of the AMV DA for short‐term severe weather forecast is assessed with three high‐impact weather events that occurred in spring 2018 and 2019 over the Great Plains of the United States. The results show that the wind and equivalent potential temperature fields associated with the storm environment and the nearby ongoing convection are improved by the AMV DA, which yields better simulation of the boundaries and the subsequent forecasts of storm evolution. For the quasi‐linear or mesoscale convective system, the assimilation of AMVs has a positive impact on the 0–3 h forecasts of composite reflectivity and accumulated precipitation in terms of the shape, location, and magnitude. However, the AMV DA has difficulty in capturing the sharp moisture gradient associated with the dryline and mostly underpredicts the associated scattered storms.
Abstract Quantitative precipitation estimation (QPE) products from the next-generation National Mosaic and QPE system (Q2) are cross-compared to the operational, radar-only product of the National Weather Service (Stage II) using the gauge-adjusted and manual quality-controlled product (Stage IV) as a reference. The evaluation takes place over the entire conterminous United States (CONUS) from December 2009 to November 2010. The annual comparison of daily Stage II precipitation to the radar-only Q2Rad product indicates that both have small systematic biases (absolute values > 8%), but the random errors with Stage II are much greater, as noted with a root-mean-squared difference of 4.5 mm day−1 compared to 1.1 mm day−1 with Q2Rad and a lower correlation coefficient (0.20 compared to 0.73). The Q2 logic of identifying precipitation types as being convective, stratiform, or tropical at each grid point and applying differential Z–R equations has been successful in removing regional biases (i.e., overestimated rainfall from Stage II east of the Appalachians) and greatly diminishes seasonal bias patterns that were found with Stage II. Biases and radar artifacts along the coastal mountain and intermountain chains were not mitigated with rain gauge adjustment and thus require new approaches by the community. The evaluation identifies a wet bias by Q2Rad in the central plains and the South and then introduces intermediate products to explain it. Finally, this study provides estimates of uncertainty using the radar quality index product for both Q2Rad and the gauge-corrected Q2RadGC daily precipitation products. This error quantification should be useful to the satellite QPE community who use Q2 products as a reference.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)