Variation in Short-term Temperature Fluctuations Across China During the Past 60 Years

Short-term temperature fluctuations(STFs),including amplitude and frequency fluctuations,are one of the main features of weather and play vital roles in determining the type of ecosystem present.Although temperature fluctuations at different time scales have been extensively discussed,the research on week-scale STFs is lacking.In this study,we developed a method,that can quantify the amplitude and frequency of STFs by the thresholds from all years.We used this method to quantify the amplitude and frequency of the 7-d STFs from 1951 to 2019 across China.Our results indicate that the amplitude of the STF was much higher in the eastern part of China than in the western part,while the frequency of the STF was higher in the middle part than in the southern and northern parts;further-more,the STF was highly dependent on internal factors such as topography.The long-term STF mainly showed a decreasing trend before 1990,which implies that temperature became increasingly stable from the 1950s to the 1990s.The main influencing factors were related to topography since the trends were relatively consistent in space.A case study in Taihu Lake showed that an unstable STF in winter and summer resulted in a smaller bloom area in the following spring and autumn.Our method could eliminate seasonal effects and is capable of analyzing STFs at scales ranging from days to years.Quantifications of the amplitude and frequency also make the STF indicators more comprehensive.Furthermore,the STF increased significantly across most of China after 1990,which implies that temperature is becoming increasingly unstable.The drivers of these STFs are related to human impacts since the trends are different in space.

Tea Plantation Frost Damage Early Warning Using a Two-Fold Method for Temperature Prediction

As the source and main producing area of tea in the world, China has formed unique tea culture, and achievedremarkable economic benefits. However, frequent meteorological disasters, particularly low temperature frostdamage in late spring has seriously threatened the growth status of tea trees and caused quality and yield reduction of tea industry. Thus, timely and accurate early warning of frost damage occurrence in specific tea garden isvery important for tea plantation management and economic values. Aiming at the problems existing in currentmeteorological disaster forecasting methods, such as difficulty in obtaining massive meteorological data, largeamount of calculation for predicted models and incomplete information on frost damage occurrence, this paperproposed a two-fold algorithm for short-term and real-time prediction of temperature using field environmentaldata, and temperature trend results from a nearest local weather station for accurate frost damage occurrence leveldetermination, so as to achieve a specific tea garden frost damage occurrence prediction in a microclimate. Timeseries meteorological data collected from a small weather station was used for testing and parameterization of atwo-fold method, and another dataset acquired from Tea Experimental Base of Zhejiang University was furtherused to validate the capability of a two-fold model for frost damage forecasting. Results showed that comparedwith the results of autoregressive integrated moving average (ARIMA) and multiple linear regression (MLR),the proposed two-fold method using a second order Furrier fitting model and a K-Nearest Neighbor model(K = 3) with three days historical temperature data exhibited excellent accuracy for frost damage occurrence prediction on consideration of both model accuracy and computation (98.46% forecasted duration of frost damage,and 95.38% for forecasted temperature at the onset time). For field test in a tea garden, the proposed methodaccurately predicted three times frost damage occurrences, including onset time, duration and occurrence level.These results suggested the newly-proposed two-fold method was suitable for tea plantation frost damage occurrence forecasting.

A TMA-Seq2seq Network for Multi-Factor Time Series Sea Surface Temperature Prediction

Sea surface temperature (SST) is closely related to global climatechange, ocean ecosystem, and ocean disaster. Accurate prediction of SST isan urgent and challenging task. With a vast amount of ocean monitoring dataare continually collected, data-driven methods for SST time-series predictionshow promising results. However, they are limited by neglecting complexinteractions between SST and other ocean environmental factors, such as airtemperature and wind speed. This paper uses multi-factor time series SSTdata to propose a sequence-to-sequence network with two-module attention(TMA-Seq2seq) for long-term time series SST prediction. Specifically, TMASeq2seq is an LSTM-based encoder-decoder architecture facilitated by factorand temporal-attention modules and the input of multi-factor time series. Ittakes six-factor time series as the input, namely air temperature, air pressure,wind speed, wind direction, SST, and SST anomaly (SSTA). A factor attentionmodule is first designed to adaptively learn the effect of different factors onSST, followed by an encoder to extract factor-attention weighted features asfeature representations. And then, a temporal attention module is designedto adaptively select the hidden states of the encoder across all time steps tolearn more robust temporal relationships. The decoder follows the temporalattention module to decode the feature vector concatenated from the weightedfeatures and original input feature. Finally, we use a fully-connect layer tomap the feature into prediction results. With the two attention modules, ourmodel effectively improves the prediction accuracy of SST since it can notonly extract relevant factor features but also boost the long-term dependency.Extensive experiments on the datasets of China Coastal Sites (CCS) demonstrate that our proposed model outperforms other methods, reaching 98.29%in prediction accuracy (PACC) and 0.34 in root mean square error (RMSE).Moreover, SST prediction experiments in China’s East, South, and Yellow Seasite data show that the proposed model has strong robustness and multi-siteapplicability.

Short-time Forecast Method of Winter Minimum Temperature in the Northern Area of Fujian

[Objective] The research aimed to study the short-time forecast method of winterminimum temperature in the northern area of Fujian.[Method] By analyzing the variation trends and distribution characteristics of extremely and averageminimum temperatures in northern Fujian in winter during 1969-2008,the relative meteorological factors which affected the low temperature weather in winter were found.The influences of relative meteorological factors on winterminimum temperature and the forecast method were summarized by combining with the climate characteristics in northern Fujian.[Result] Winterminimum temperature in Guangze and Pucheng in the north of northern Fujian was the lowest.The second one was in Shaowu,Wuyishan,Jianyang,Songxi and Zhenghe.Theminimum temperature in Jian’ou and Shunchang was higher and was the highest in Yanping.Theminimum temperature mainly depended on the temperature reduction degree from the afternoon to the night.The temperature reduction degree varied with the sky condition and cold air intensity.The temperature reduction included the advection,radiation,advection-radiation and non-advection-radiation types.The temperature had the different reduction characteristics under the different sky conditions.The forecast ofminimum temperature should be carried out based on the weather typing.Meanwhile,the successful forecast key ofminimum temperature was grasping the shift pathway and speed of cold air.[Conclusion] The research provided the theory basis for improving the forecast accuracy of winterminimum temperature.

Efficient and robust CNN-LSTM prediction of flame temperature aided light field online tomography

Light field tomography,an optical combustion diagnostic technology,has recently attracted extensive attention due to its easy implementation and non-intrusion.However,the conventional iterative methods are high data throughput,low efficiency and time-consuming,and the existing machine learning models use the radiation spectrum information of the flame to realize the parameter field measurement at the current time.It is still an offline measurement and cannot realize the online prediction of the instantaneous structure of the actual turbulent combustion field.In this work,a novel online prediction model of flame temperature instantaneous structure based on deep convolutional neural network and long short-term memory(CNN-LSTM)is proposed.The method uses the characteristics of local perception,shared weight,and pooling of CNN to extract the threedimensional(3D)features of flame temperature and outgoing radiation images.Moreover,the LSTM is used to comprehensively utilize the ten historical time series information of high dynamic combustion flame to accurately predict 3D temperature at three future moments.A chaotic time-series dataset based on the flame radiation forward model is built to train and validate the performance of the proposed CNN-LSTM model.It is proven that the CNN-LSTM prediction model can successfully learn the evolution pattern of combustion flame and make accurate predictions.

The characteristics of temperature variability with terrain, latitude and longitude in Sichuan- Chongqing Region

Using the daily temperature data of 95 meteorological stations from Sichuan-Chongqing Region and its surrounding areas, this paper adopted these methods （e.g., linear regression, trend coefficient, geographical statistics, gray relational analysis and spatial analysis functions of GIS） to analyze the relations of temperature variability with topography, latitude and longitude. Moreover, the rank of gray correlation between temperature variability and elevation, longitude, latitude, topographic position and surface roughness also was meas- ured. These results indicated： （1） The elevation affected temperature variability most obviously, followed by latitude, and longitude. The slope of the linear regression between temperature change rate and elevation, latitude and longitude was 0.4142, 0.0293 and -0.3270, respectively （2） The rank of gray correlation between temperature change rate and geographic factors was elevation 〉 latitude 〉 surface roughness 〉 topographic position 〉 longitude. The gray correla- tion degree between temperature change rate and elevation was 0.865, followed by latitude with 0.796, and longitude with 0.671. （3） The rate of temperature change enhanced with the increase of elevation. Especially, the warming trend was significant in the plateau and mountain areas of western Sichuan, and mountain and valley areas of southwestern Sichuan （with the warming rate of 0.74℃/10a during the 1990s）. However, there was a weak warming trend in Sichuan Basin and its surrounding low mountain and hilly areas. （4） The effects of latitude on temperature change rate presented the specific regulation, which the warming rate of low-latitude areas was more significant than that of high-latitude areas. However, they were consistent with the regulation that the increasing of low temperature controlled most of the warming trend, due to the effects of terrain and sically, temperature variability along longitude elevation on annual mean temperature. （5） Ba- direction resulted from the regular change of elevation along longitude. It was suggested that, in Sichuan-Chongqing Region, special features of temperature variability largely depended on the terrain complexity （e.g., undulations, mutations and roughness）. The elevation level controlled only high or low annual mean temperature and the range of temperature change rate in the macro sense.

Li-ion battery temperature estimation based on recurrent neural networks

The monitoring of Li-ion battery temperatures is essential to ensure high efficiency and safety.In this work,two types of recurrent neural networks (RNNs),which are long short-term memory-RNN (LSTM-RNN) and gated recurrent unit-RNN(GRU-RNN),are proposed to estimate the surface temperature of 18650 Li-ion batteries during the discharging processes under different ambient temperatures.The datasets acquired from the Prognostics Center of Excellence (PCo E) of NASA are used to train,validate and test the networks.In previous work,temperature has been set as the output of the networks;however,here the temperature difference along the time axis is adopted as the output.The net heat generated results in net temperature change,which is more closely aligned with electrochemical and thermodynamic laws.Extensive simulation results show that the two RNNs can achieve accurate real-time battery temperature estimation.The maximum absolute error in temperature estimation is approximately 0.75°C and the correlation coefficient between the estimated and measured temperature curves is greater than 0.95.The influences of three crucial parameters,which are the number of hidden neurons,initial learning rate and maximum number of iterations,are also assessed in terms of training time,root mean square error and mean absolute error.

Summertime Atmospheric Teleconnection Pattern Associated with a Warming over the Eastern Tibetan Plateau

By using a surface air temperature index （SATI） averaged over the eastern Tibetan Plateau （TP）, investigation is conducted on the short-term climate variation associated with the interannual air warming （or cooling） over the TP in each summer month. Evidence suggests that the SATI is associated with a consistent teleconnection pattern extending from the TP to central-western Asia and southeastern Europe. Associated rainfall changes include, for a warming case, a drought in northern India in May and June, and a stronger mei-yu front in June. The latter is due to an intensified upper-level northeasterly in eastern China and a wetter and warmer condition over the eastern TP. In the East Asian regions, the time-space distributions of the correlation patterns between SATI and rainfall are more complex and exhibit large differences from month to month. Some studies have revealed a close relationship between the anomalous heating over the TP and the rainfall anomaly along the Yangtze River valley appearing in the summer on a seasonal mean time-scale, whereas in the present study, this relationship only appears in June and the signal＇s significance becomes weaker after the long-term trend in the data was excluded. Close correlations between SATI and the convection activity and SST also occur in the western Pacific in July and August： A zonally-elongated warm tone in the SST in the northwestern Pacific seems to be a passive response of the associated circulation related to a warm SATI. The SATI-associated teleconnection pattern provides a scenario consistently linking the broad summer rainfall anomalies in Europe, central-western Asia, India, and East Asia.

The effects of climate warming on microbe-mediated mechanisms of sediment carbon emission

Due to significant differences in biotic and abiotic properties of soils compared to those of sediments,the predicted underlying microbe-mediated mechanisms of soil carbon emissions in response to warming may not be applicable for estimating similar emissions from inlandwater sediments.We addressed this issue by incubating different types of sediments,(including lake,small river,and pond sediments)collected from 36 sites across the Yangtze River basin,under short-term experimentalwarming to explore the effects of climatewarming on sediment carbon emission and the underlying microbe-mediated mechanisms.Our results indicated that under climate warming Cc emissions were affected more than CH_(4) emissions,and that pond sediments may yield a greater relative contribution of CO_(2) to total carbon emissions than lake and river sediments.Warming-induced CO_(2) and CH_(4) increases involve different microbe-mediated mechanisms;Warming-induced sediment CO_(2) emissions were predicted to be directly positively driven by microbial community network modularity,which was significantly negatively affected by the quality and quantity of organic carbon and warming-induced variations in dissolved oxygen,Conversely,warminginduced sediment CH_(4) emissionswere predicted to be directly positively driven bymicrobial community network complexity,which was significantly negatively affected by warminginduced variations in pH.Our findings suggest that biotic and abiotic drivers for sediment CO_(2) and CH_(4) emissions in response to climate warming should be considered separately when predicting sediment organic carbon decomposition dynamics resulting from climate change.