Forecasts Of Surface Air Temperature Research Articles

Pattern Projection Calibrations on Subseasonal Forecasts of Surface Air Temperature over East Asia

Abstract Subseasonal forecasts have recently attracted widespread interest yet remain a challenging issue. A statistical Kalman filter pattern projection method (KFPPM), which absorbs the projection conception of the raw covariance pattern projection (COVPPM) and the adaptive adjustments of the Kalman filter, is proposed to calibrate the single-model forecasts of the daily maximum and minimum temperatures (Tmax and Tmin) for lead times of 8–42 days over East Asia in 2018 derived from the UKMO control (CTL) forecast. The Kalman filter–based gridly calibration (KFGC) is carried out in parallel as a benchmark, which could improve the forecast skills to a certain extent. The COVPPM effectively calibrates the temperature forecasts at the early stage and displays better performances than the CTL and KFGC. However, with the growing lead times, it shows speedily decreasing skills and can no longer produce positive adjustments over the areas outside the plateaus. By contrast, the KFPPM consistently outperforms the other calibrations and reduces the forecast errors by almost 1.0° and 0.5°C for Tmax and Tmin, respectively, both retaining superiorities to the random climatology benchmark till the lead time of 24 days. The optimization of KFPPM maintains throughout the whole range of the subseasonal time scale, showing the most conspicuous improvements distributed over the Tibetan Plateau and its surroundings. Though the postprocessing procedures are more skillful in calibrating Tmax forecasts than Tmin forecasts, the Tmax forecasts are still characterized by lower skills than the latter. Case experiments further demonstrate the abovementioned features and imply the potential capability of KFPPM in improving forecast skills and disaster preventions for extreme temperature events.

Forecast calibrations of surface air temperature over Xinjiang based on U-net neural network

In this study, a deep learning method named U-net neural network is utilized to calibrate the gridded forecast of surface air temperature from the Global Ensemble Forecasting System (GEFS), with forecast lead times of 1–7 days in Xinjiang. The calibration performance of U-net is compared with three conventional postprocessing methods: unary linear regression (ULR), the decaying averaging method (DAM) and Quantile Mapping (QM). Results show that biases of the raw GEFS forecasts are mainly distributed in the Altai Mountains, the Junggar Basin, the Tarim Basin and the Kunlun Mountains. The four postprocessing methods effectively improve the forecast skills for all lead times, whereas U-net shows the best correction performance with the lowest mean absolute error (MAE) and the highest hit rate of 2°C (HR2) and pattern correlation coefficient (PCC). The U-net model considerably reduces the warm biases of the raw forecasts. The skill improvement magnitudes are greater in southern than northern Xinjiang, showing a higher mean absolute error skill score (MAESS). Furthermore, in order to distinguish the error sources of each forecasting scheme and to reveal their capabilities of calibrating errors of different sources, the error decomposition analysis is carried out based on the mean square errors. It shows that the bias term is the leading source of error in the raw forecasts, and barely changes as the lead time increases, which is mainly distributed in Tarim Basin and Kunlun Mountains. All four forecast calibrations effectively reduce the bias and distribution error of the raw forecasts, but only the U-net significantly reduces the sequence error.

Оценки влияния подстилающей поверхности на точность численного прогноза температуры воздуха на территории Беларуси с использованием модели WRF

The spatial structure of land use and biophysical characteristics of land surface (albedo, leaf index, and vegetation cover) are updated using the GLASS (Global Land Surface Satellite) and GLC2019 (Global Land Cover, 2019) modern satellite databases for mesoscale numerical weather prediction with the WRF model for the territory of Belarus. The series of WRF-based numerical experiments was performed to verify the influence of the updated characteristics on the forecast quality for some difficult to predict winter cases. The model was initialized by the GFS (Global Forecast System, NCEP) global numerical weather prediction model. It is shown that the use of high-resolution land use data in the WRF and the consideration of the new albedo and leaf index distribution over the territory of Belarus can reduce the root-mean-square error (RMSE) of short-range (to 48 hours) forecasts of surface air temperature by 16–33% as compared to the GFS. The RMSE of the temperature forecast for the weather stations in Belarus for a forecast lead time of 12, 24, 36, and 48 hours decreased on average by 0.40°С (19%), 0.35°С (10%), 0.68°С (23%), and 0.56°С (15%), respectively. The most significant decrease in RMSE of the numerical forecast of temperature (up to 2.1 °С) was obtained for the daytime (for a lead time of 12 and 36 hours), when positive feedbacks between albedo and temperature of the land surface are manifested most. Keywords: numerical weather prediction, WRF, digital land surface model, albedo, leaf area index, forecast model validation

Statistical Calibrations of Surface Air Temperature Forecasts over East Asia using Pattern Projection Methods

AbstractIn this study, two pattern projection methods, i.e., the Stepwise Pattern Projection Method (SPPM) and the newly proposed Neighborhood Pattern Projection Method (NPPM), are investigated to improve forecast skills of daily maximum and minimum temperatures (Tmax and Tmin) over East Asia with lead times of 1–7 days. Meanwhile, the decaying averaging method (DAM) is conducted in parallel for comparison. These post-processing methods are found to effectively calibrate the temperature forecasts on the basis of the raw ECMWF output. Generally, the SPPM is slightly inferior to the DAM, while its insufficiency decreases with increasing lead times. The NPPM shows manifest superiority for all lead times, with the mean absolute errors of Tmax and Tmin decreased by ~0.7°C and ~0.9°C, respectively. Advantages of the two pattern projection methods are both mainly concentrated on the high-altitude areas such as the Tibetan Plateau, where the raw ECMWF forecasts show most conspicuous biases. In addition, aiming at further assessments of these methods on extreme event forecasts, two case experiments are carried out towards a heat wave and a cold surge, respectively. The NPPM is retained as the optimal with the highest forecast skills, which reduces most of the biases to < 2°.C for both Tmax and Tmin over all the lead days. In general, the statistical pattern projection methods are capable of effectively eliminating spatial biases in forecasts of surface air temperature. Compared with the initial SPPM, the NPPM not only produces more powerful forecast calibrations, but also provides more pragmatic calculations and greater potential economic benefits in practical applications.

Климатические особенности зимнего сезона 2020/2021 гг. и прогностические оценки температурно-влажностного режима на лето 2021 г. по территории Северной Евразии

Climatic features of the 2020/2021 winter season and the air temperature and precipitation outlook for the summer of 2021 over Northern Eurasia / Khan V.M., Vilfand R.M., Emelina E.V., Kaverina E.S., Kulikova I.A., Sumerova K.A., Tischenko V.A. // Hydrometeorological Research and Forecasting, 2021, no. 2 (380), pp. 6-19. The main features of the Northern Hemisphere large-scale atmospheric circulation are analyzed for the past 2020/2021 winter. The accuracy of consensus forecasts of air temperature and precipitation compiled during the work of the 19th session of the North Eurasian Climate Outlook Forum (NEACOF-19) is presented, with the skill scores of consensus forecasts for Northern Eurasia. The main features of the thermal state of the ocean and large-scale atmospheric circulation for the coming summer of 2021 are considered and analyzed. A forecast of surface air temperature and precipitation anomalies for the summer of 2021 agreed with the NEACOF-20 experts is formulated. Keywords: North Eurasian Climate Outlook Forum, North Eurasian Climate Center, consensus forecast, air temperature, precipitation, large-scale atmospheric circulation, hydrodynamic models, sea surface temperature

Опыт использования фильтра Калмана для коррекции численного прогноза приземной температуры воздуха

A possibility of using the one-dimensional Kalman filter to improve the forecast of surface air temperature at an irregular grid of point is studied. This mechanism is tested using the forecasts obtained from different configurations of two different numerical weather prediction models. An algorithm for the statistical correction of numerical forecasts of surface air temperature based on the one-dimensional Kalman filter is constructed. Two methods are proposed for estimating the bias noise dispersion. The series of experiments demonstrated the effectiveness of the algorithm for the bias compensation.The most significantresults are achieved for the models with large bias or for long-range forecasts. At the same time, the use of the algorithm has little effect on the root-meansquare error of the forecast. Keywords: hydrodynamic model of the atmosphere, numerical weather prediction, statistical correction of numerical forecasts, Kalman filter

Subseasonal Forecast of Surface Air Temperature Using Superensemble Approaches: Experiments over Northeast Asia for 2018

AbstractBridging the gap between weather forecasting and climate prediction, subseasonal to seasonal (S2S) forecasts are of great importance yet currently of relatively poor quality. Using the S2S Prediction Project database, the study evaluates products derived from four operational centers of CMA, KMA, NCEP, and UKMO, and superensemble experiments including the straightforward ensemble mean (EMN), bias-removed ensemble mean (BREM), error-based superensemble (ESUP), and Kalman filter superensemble (KF), in forecasts of surface air temperature with lead times of 6–30 days over northeast Asia in 2018. Validations after the preprocessing of a 5-day running mean suggest that the KMA model shows the highest skill for either the control run or the ensemble mean. The nonequal weighted ESUP is slightly superior to BREM, whereas they both show larger biases than EMN after a lead time of 22 days. The KF forecast constantly outperforms the others, decreasing mean absolute errors by 0.2°–0.5°C relative to EMN. Forecast experiments of the 2018 northeast Asia heat wave reveal that the superensembles remarkably improve the raw forecasts featuring biases of >4°C. The prominent advancement of KF is further confirmed, showing the regionally averaged bias of ≤2°C and the hit rate of 2°C reaching up to 60% at a lead time of 22 days. The superensemble techniques, particularly the KF method of dynamically adjusting the weights in accordance with the latest information available, are capable of improving forecasts of spatiotemporal patterns of surface air temperature on the subseasonal time scale, which could extend the skillful prediction lead time of extreme events such as heat waves to about 3 weeks.

The Met Office Operational Soil Moisture Analysis System

In this study, the current Met Office operational land surface data assimilation system used to produce soil moisture analyses is presented. The main aim of including Land Surface Data Assimilation (LSDA) in both the global and regional systems is to improve forecasts of surface air temperature and humidity. Results from trials assimilating pseudo-observations of 1.5 m air temperature and specific humidity and satellite-derived soil wetness (ASCAT) observations are analysed. The pre-processing of all the observations is described, including the definition and construction of the pseudo-observations. The benefits of using both observations together to produce improved forecasts of surface air temperature and humidity are outlined both in the winter and summer seasons. The benefits of using active LSDA are quantified by the root mean squared error, which is computed using both surface observations and European Centre for Medium-Range Weather Forecasts (ECMWF) analyses as truth. For the global model trials, results are presented separately for the Northern (NH) and Southern (SH) hemispheres. When compared against ground-truth, LSDA in winter NH appears neutral, but in the SH it is the assimilation of ASCAT that contributes to approximately a 2% improvement in temperatures at lead times beyond 48 h. In NH summer, the ASCAT soil wetness observations degrade the forecasts against observations by about 1%, but including the screen level pseudo-observations provides a compensating benefit. In contrast, in the SH, the positive effect comes from including the ASCAT soil wetness observations, and when both observations types are assimilated there is a compensating effect. Finally, we demonstrate substantial improvements to hydrological prediction when using land surface data assimilation in the regional model. Using the Nash-Sutcliffe Efficiency (NSE) metric as an aggregated measure of river flow simulation skill relative to observations, we find that NSE was improved at 106 of 143 UK river gauge locations considered after LSDA was introduced. The number of gauge comparisons where NSE exceeded 0.5 is also increased from 17 to 28 with LSDA.

Statistical Correction of the COSMO Model Weather Forecasts Based on Neural Networks

Different methods for the statistical correction of the forecasts of surface parameters using the COSMO-Ru13-ENA model with the lead time up to 117 hours are considered. The methods include the systematic correction using the data from recent observations at a weather station, the correction based on special neural networks as well as different combinations of these two techniques. The study presents the estimates of the results of applying the analyzed correction methods to the forecasts of surface air temperature, dew point, and wind speed modulus based on the independent sample for 2018 with the total volume of 2.34 × 107 forecasts. The correction method based on neural networks reduces forecast errors even at the points where meteorological observations have not been carried out.

How Much Can AI Techniques Improve Surface Air Temperature Forecast? —A Report from AI Challenger 2018 Global Weather Forecast Contest

In August 2018, the Institute of Urban Meteorology (IUM) in Beijing co-organized with Sinovation Ventures a Weather Forecasting Contest (WFC)—one of the AI (artificial intelligence) Challenger Global Contests. The WFC aims to take advantage of the AI techniques to improve the quality of weather forecast. Across the world, more than 1000 teams enrolled in the WFC and about 250 teams completed real-time weather forecasts, among which top 5 teams were awarded in the final contest. The contest results show that the AI-based ensemble models exhibited improved skill for forecasts of surface air temperature and relative humidity at 2-m and wind speed at 10-m height. Compared to the IUM operational analog ensemble weather model forecast, the most notable improvements of 24.2% and 17.0% in forecast accuracy for surface 2-m air temperature are achieved by two teams using the AI techniques of time series model, gradient boosting tree, depth probability prediction, and so on. Meanwhile, it is found that reasonable data processing techniques and model composite structure are also important for obtaining better forecasts.

基于BMA方法的地面气温概率预报研究

基于BMA方法的地面气温概率预报研究

Operative Scheme for the Short-range Complex Forecasting of Surface Air Temperature and Humidity

The statistical scheme is proposed for the forecast of surface air temperature and humidity using operative weather forecasts with 3–5-day lead time from the best forecasting hydrodynamic models as well as the archives of forecasts of these models and observational data from 2800 weather stations of Russia, Eastern Europe, and Central Asia. The output of the scheme includes the forecasts of air temperature for the standard observation moments with the period of 6 hours and extreme temperatures with the lead times of 12–120 hours. The accuracy of temperature and humidity forecasts for the period from July 2014 till June 2017 is much higher than that for the forecasts of original hydrodynamic models. The skill scores for extreme temperature forecasts based on the proposed method are compared with the similar results of the Weather Element Computation (WEC) forecasting scheme and forecasts by weathermen.

Impact of Model Physics on Seasonal Forecasts of Surface Air Temperature in the Arctic

The impacts of model physics and initial sea ice thickness on seasonal forecasts of surface energy budget and air temperature in the Arctic during summer were investigated based on Climate Forecast System, version 2 (CFSv2), simulations. The model physics changes include the enabling of a marine stratus cloud scheme and the removal of the artificial upper limit on the bottom heat flux from ocean to sea ice. The impact of initial sea ice thickness was examined by initializing the model with relatively realistic sea ice thickness generated by the Pan-Arctic Ice Ocean Modeling and Assimilation System (PIOMAS). Model outputs were compared to that from a control run that did not impose physics changes and used Climate Forecast System Reanalysis (CFSR) sea ice thickness. After applying the physics modification to either sea ice thickness initialization, the simulated total cloud cover more closely resembled the observed monthly variations of total cloud cover except for the midsummer reduction. Over the Chukchi–Bering Seas, the model physics modification reduced the seasonal forecast bias in surface air temperature by 24%. However, the use of initial PIOMAS sea ice thickness alone worsened the surface air temperature predictions. The experiment with physics modifications and initial PIOMAS sea ice thickness achieves the best surface air temperature improvement over the Chukchi–Bering Seas where the area-weighted forecast bias was reduced by 71% from 1.05 K down to −0.3 K compared with the control run. This study supports other results that surface temperatures and sea ice characteristics are highly sensitive to the Arctic cloud and radiation formulations in models and need priority in model formulation and validation.

English

The determination of variance of surface air temperature is very essential since it has a direct impact on vegetation, environment and human livelihood. Forecast of surface air temperature is difficult because of the complex physical phenomenon and the random-like behavior of atmospheric system which influences the temperature event on the earth surface. In this study, forecast models based on artificial neural network (ANN) and genetic programming (GP) approaches were proposed to predict lead seven days minimum and maximum surface air temperature using the weather parameters observed at the station Chennai, India. The outcome of this study stated that models formulated using ANN approach are more accurate than genetic programming for all seven days with the highest coefficient of determination (R2), least mean absolute error (MAE), root mean square error (RMSE) and mean absolute percentage error (MAPE) on deployment with independent test dataset. ANN models give statistically acceptable mean absolute error of 0.59oC for lead day one in minimum temperature forecast and 0.86oC variance for lead day one in maximum temperature forecast. The study also clarified that the level of accuracy of the proposed prediction models were found to be better for smaller lead days when compared with higher lead days with both approaches.

Dynamical Downscaling of Austral Summer Climate Forecasts over Southern Africa Using a Regional Coupled Model

Abstract The prediction skill of dynamical downscaling is evaluated for climate forecasts over southern Africa using the Advanced Research Weather Research and Forecasting (WRF) model. As a case study, forecasts for the December–February (DJF) season of 2011/12 are evaluated. Initial and boundary conditions for the WRF model were taken from the seasonal forecasts of the Scale Interaction Experiment-Frontier Research Center for Global Change (SINTEX-F) coupled general circulation model. In addition to sea surface temperature (SST) forecasts generated by nine-member ensemble forecasts of SINTEX-F, the WRF was also configured to use SST generated by a simple mixed layer Price–Weller–Pinkel ocean model coupled to the WRF model. Analysis of the ensemble mean shows that the uncoupled WRF model significantly increases the biases (errors) in precipitation forecasted by SINTEX-F. When coupled to a simple mixed layer ocean model, the WRF model improves the spatial distribution of precipitation over southern Africa through a better representation of the moisture fluxes. Precipitation anomalies forecasted by the coupled WRF are seen to be significantly correlated with the observed precipitation anomalies over South Africa, Zimbabwe, southern Madagascar, and parts of Zambia and Angola. This is in contrast to the SINTEX-F global model precipitation anomaly forecasts that are closer to observations only for parts of Zimbabwe and South Africa. Therefore, the dynamical downscaling with the coupled WRF adds value to the SINTEX-F precipitation forecasts over southern Africa. However, the WRF model yields positive biases (>2°C) in surface air temperature forecasts over the southern African landmass in both the coupled and uncoupled configurations because of biases in the net heat fluxes.

Prospects of using Bayesian model averaging for the calibration of one-month forecasts of surface air temperature over South Korea

In this study, we investigated the prospect of calibrating probabilistic forecasts of surface air temperature (SAT) over South Korea by using Bayesian model averaging (BMA). We used 63 months of simulation results from four regional climate models (RCMs) with two boundary conditions (NCEP-DOE and ERA-interim) over the CORDEX East Asia. Rank histograms and residual quantile-quantile (R-Q-Q) plots showed that the simulation skills of the RCMs differ according to season and geographic location, but the RCMs show a systematic cold bias irrespective of season and geographic location. As a result, the BMA weights are clearly dependent on geographic location, season, and correlations among the models. The one-month equal weighted ensemble (EWE) outputs for the 59 stations over South Korea were calibrated using the BMA method for 48 monthly time periods based on BMA weights obtained from the previous 15 months of training data. The predictive density function was calibrated using BMA and the individual forecasts were weighted according to their performance. The raw ensemble forecasts were assessed using the flatness of the rank histogram and the R-Q-Q plot. The results showed that BMA improves the calibration of the EWE and the other weighted ensemble forecasts irrespective of season, simulation skill of the RCM, and geographic location. In addition, deterministic-style BMA forecasts usually perform better than the deterministic forecast of the single best member.

Impacts of Snow Initialization on Subseasonal Forecasts of Surface Air Temperature for the Cold Season

Abstract The present study examines the impacts of snow initialization on surface air temperature by a number of ensemble seasonal predictability experiments using the NCAR Community Atmosphere Model version 3 (CAM3) AGCM with and without snow initialization. The study attempts to isolate snow signals on surface air temperature. In this preliminary study, any effects of variations in sea ice extent are ignored and do not explicitly identify possible impacts on atmospheric circulation. The Canadian Meteorological Center (CMC) daily snow depth analysis was used in defining initial snow states, where anomaly rescaling was applied in order to account for the systematic bias of the CAM3 snow depth with respect to the CMC analysis. Two suites of seasonal (3 months long) ensemble hindcasts starting at each month in the colder part of the year (September–April) with and without the snow initialization were performed for 12 recent years (1999–2010), and the predictability skill of surface air temperature was estimated. Results show that considerable potential predictability increases up to 2 months ahead can be attained using snow initialization. Relatively large increases are found over East Asia, western Russia, and western Canada in the later part of this period. It is suggested that the predictability increases are sensitive to the strength of snow–albedo feedback determined by given local climate conditions; large gains tend to exist over the regions of strong snow–albedo feedback. Implications of these results for seasonal predictability over the extratropical Northern Hemisphere and future direction for this research are discussed.

겨울철 동아시아 지역 기온의 계절 예측에 눈깊이 초기화가 미치는 영향

Does snow depth initialization have a quantitative impact on sub-seasonal to seasonal prediction skill? To answer this question, a snow depth initialization technique for seasonal forecast system has been implemented and the impact of the initialization on the seasonal forecast of surface air temperature during the wintertime is examined. Since the snow depth observation can not be directly used in the model simulation due to the large systematic bias and much smaller model variability, an anomaly rescaling method to the snow depth initialization is applied. Snow depth in the model is initialized by adding a rescaled snow depth observation anomaly to the model snow depth climatology. A suite of seasonal forecast is performed for each year in recent 12 years (1999-2010) with and without the snow depth initialization to evaluate the performance of the developed technique. The results show that the seasonal forecast of surface air temperature over East Asian region sensitively depends on the initial snow depth anomaly over the region. However, the sensitivity shows large differences for different timing of the initialization and forecast lead time. Especially, the snow depth anomaly initialized in the late winter (Mar. 1) is the most effective in modulating the surface air temperature anomaly after one month. The real predictability gained by the snow depth initialization is also examined from the comparison with observation. The gain of the real predictability is generally small except for the forecasting experiment in the early winter (Nov. 1), which shows some skillful forecasts. Implications of these results and future directions for further development are discussed.

Bias Correction and Forecast Skill of NCEP GFS Ensemble Week-1 and Week-2 Precipitation, 2-m Surface Air Temperature, and Soil Moisture Forecasts

Abstract A simple bias correction method was used to correct daily operational ensemble week-1 and week-2 precipitation and 2-m surface air temperature forecasts from the NCEP Global Forecast System (GFS). The study shows some unexpected and striking features of the forecast errors or biases of both precipitation and 2-m surface air temperature from the GFS. They are dominated by relatively large-scale spatial patterns and low-frequency variations that resemble the annual cycle. A large portion of these forecast errors is removable, but the effectiveness is time and space dependent. The bias-corrected week-1 and week-2 ensemble precipitation and 2-m surface air temperature forecasts indicate some improvements over their raw counterparts. However, the overall levels of week-1 and week-2 forecast skill in terms of spatial anomaly correlation and root-mean-square error are still only modest. The dynamical soil moisture forecasts (i.e., land surface hydrological model forced with bias-corrected precipitation and 2-m surface air temperature integrated forward for up to 2 weeks) have very high skill, but hardly beat persistence over the United States. The inability to outperform persistence mainly relates to the skill of the current GFS week-1 and week-2 precipitation forecasts not being above a threshold (i.e., anomaly correlation > 0.5 is required).

Submonthly Forecasting of Winter Surface Air Temperature in North America Based on Organized Tropical Convection

Using outgoing longwave radiation (OLR) data as a proxy for tropical convection, the relationship between the Madden-Julian Oscillation (MJO)-related tropical convection and North American winter surface air temperature (SAT) is investigated. A lagged regression analysis between one of the leading OLR principal components (PCs) and SAT shows highly significant correlations over south-southeastern Canada, with the strongest SAT anomalies occurring at a lag of 2 to 3 pentads. The lagged temperature composite shows an extensive intensified positive SAT anomaly over southern Canada and the northern United States 10–15 days after the occurrence of enhanced convection over the Indian Ocean and the Maritime Continent and reduced convection near the tropical central Pacific. The lagged regressions of 500 hPa geopotential height with the OLR PC reveal a wave train propagating from the tropics to North America. This observation has been reproduced in a simple general circulation model (SGCM) with an idealized forcing that mimics the tropical MJO convection anomaly. A simple regression model is constructed to predict winter SAT over North America with tropical MJO convective activity as the single predictor. Under a cross-validation framework, this statistical model produces 10–15 day SAT forecasts with modest, yet statistically significant, skill in winter over North America.