111 results about "Climate model" patented technology

The invention relates to the technical field of data reasoning, and discloses a regional medium-and-long-term new energy power prediction method, which comprises the following steps: acquiring historical new energy output time sequence data and corresponding meteorological element data of a target region, and generating a regional basic data set; analyzing space-time coupling characteristics in the regional basic data set, and extracting a cross-regional new energy output collaborative fluctuation mode in the space-time coupling characteristics; based on the cross-regional new energy output collaborative fluctuation mode, constructing an output characteristic migration matrix of the target region; fusing the output characteristic migration matrix and climate mode forecast data to obtain a medium and long term weather-output mapping relation of the target area; and outputting a new energy power prediction result of the target area according to the medium and long term weather-output mapping relation, the method can improve the accuracy of medium and long term new energy power prediction.

The invention relates to the field of ecological environment and climate change research, and discloses a vegetation phenology change trend comprehensive evaluation method, which comprises the following steps: acquiring climate data and tree ring data of a target area, simulating a relationship between radial growth and phenology of a tree by using a Vaganov-Shashki n model based on the acquired data, and generating a time sequence of phenology change; extracting phenological key nodes such as growing season starting time, end time and growing season length; analyzing the driving effect of climatic factors on phenological key nodes, and determining main influence factors; on the basis of the climate model and different emission scenes, predicting the vegetation phenological trend under future climate change; the response of the regional ecosystem to climate changes is evaluated based on historical and future data. Through simulation and quantitative analysis, a scientific basis is provided for long-term trend prediction and regional ecological management of vegetation phenological change, and the method has high applicability and reliability.

The invention discloses a rapid positioning and early warning system for ambrosia trifida in forest planting, and the system comprises a climate and growth prediction module which analyzes historical climate data and the growth law of ambrosia trifida, combines with a climate model, and predicts the rapid positioning and early warning of ambrosia trifida under different climate change conditions. The growth mode, the expansion trend and the possible distribution area of ambrosia trifida are determined; the real-time monitoring and data acquisition module is used for collecting environmental data and the growth state of ambrosia trifida in real time; the data fusion and analysis module is used for comprehensively processing the multi-source data collected by the climate and growth prediction module and the real-time monitoring module, evaluating the growth trend, spatial distribution and ecological risk of ambrosia trifida and providing accurate data support for subsequent early warning and decision making; and the early warning and decision support module automatically generates early warning information of different risk levels according to the growth trend and the risk assessment result generated by the data fusion and analysis module. The method has the advantages of high positioning precision, strong adaptive adjustment capability, accurate grading early warning mechanism and the like.

The invention relates to the technical field of hydrology and water resources, in particular to a medium-and-long-term reservoir runoff intelligent forecasting technology coupling a climate mode and a bidirectional LSTM. According to the technical scheme, the method comprises the following steps of (1) constructing a multi-source heterogeneous data set, (2) automatically screening variables most related to runoff changes by adopting an MIC method and eliminating redundant features, (3) constructing an end-to-end prediction model composed of a CNN-GRU module and a bidirectional LSTM (Bi-LSTM) module, decomposing a runoff time sequence through an empirical mode decomposition method, and calculating the runoff time sequence according to the runoff time sequence. The method comprises the steps of (1) carrying out prediction on runoff and optimizing a prediction result by adopting a multi-factor nearest neighbor regression method, (2) training a model by adopting historical hydrometeorological data and evaluating prediction performance, and (3) inputting data to the trained model, obtaining a medium-and-long-term runoff prediction result, and optimizing final prediction output by combining a deviation correction method. According to the method, the precision and stability of medium and long-term runoff prediction are improved by coupling the climate mode and the deep learning method.

The invention discloses a greenhouse gas concentration time sequence prediction method based on a sudden change perception attention mechanism. The method comprises the steps of data preprocessing, sudden change intensity sequence construction with boundary processing, time sequence feature coding, sudden change perception attention weight calculation, context vector generation and concentration prediction, model training and optimization and model prediction. The method aims to solve the problem that a standard deep learning model is slow in sensing and lagged in prediction for a sudden change event in a concentration sequence, and finally realizes high-precision prediction for future concentration change, especially a sudden concentration peak value by endowing the model with the capability of actively identifying and reinforcing the learning of a historical sudden change mode. The urgent demand for early warning of abnormal emission in practical application is met. The method is particularly suitable for processing foundation observation data with small resolution and even higher resolution, has the core value of improving the prediction capability of concentration dramatic change driven by sudden emission events, and can be widely applied to key scenes such as accurate carbon emission monitoring, environmental pollution early warning and climate model simulation.

The invention relates to a medium and long-term runoff prediction technology, discloses a physical process model-based medium and long-term runoff prediction method for a complex terrain high-cold region, and aims to solve the problems of low precision, poor applicability and the like of medium and long-term runoff prediction in the complex terrain high-cold region in the prior art. The method comprises the following steps: S1, calibrating and considering a hydrological mode of a snow melting process aiming at a research area by utilizing a land surface fusion data set; s2, driving a mesoscale weather mode to carry out high-temporal-spatial-resolution dynamic downscaling simulation on the research area by utilizing circulation information of a future preset time period predicted by an international mainstream business dynamic climate mode, and selecting an optimal operation scheme through evaluation, predicting high temporal-spatial resolution meteorological element information on a long-term time scale in a future preset time period; and S3, taking the predicted medium-and-long-term high-temporal-spatial-resolution near-surface hydrometeorological data in the future preset time period as driving data, and performing medium-and-long-term runoff prediction on the research area in combination with the near-real-time land surface fusion data and the calibrated hydrological mode.

The invention discloses a method for predicting reservoir inflow and under-dam water resource change trend in a future change environment. The method comprises the steps of obtaining future monthly rainfall data, future monthly temperature data, future monthly evapotranspiration data and future land utilization data based on a GCM climate mode; the future runoff process of the reservoir control section is obtained through simulation in combination with an RCCC-WBM model; based on a data mining technology and an LSTM deep learning coupling model, mining relation characteristics of future reservoir runoff and the under-dam water resource quantity, and obtaining the under-dam water resource quantity by inputting the reservoir runoff; and a nonlinear trend analysis method and the like are adopted to analyze change trend characteristics of water resources under the dam in the future. The method can fully solve the problem of insufficient consideration of climate change and land utilization in the existing method, and can significantly improve the prediction precision of the water resource under the dam in the future.

The invention discloses a sub-seasonal climate objective prediction method and system based on machine learning and multi-scale circulation, and the method comprises the steps: carrying out the conversion of meteorological element data of a meteorological station in a target region, and obtaining the main modes and time coefficient sequences of climate anomalies in different prediction periods through EOF decomposition; the method comprises the following steps: intelligently screening atmosphere reanalysis circulation data and sub-seasonal climate mode circulation prediction data to establish a dynamic circulation prediction factor set; modeling and evaluating the time coefficient corresponding to the main mode of the climate anomaly by using a random forest, and constructing random forest climate objective prediction models of different prediction time periods; based on the dynamic circulation prediction factor set, taking a random forest climate objective prediction model as an initial prediction result, and utilizing a gradient boosting tree to construct a gradient boosting tree climate objective prediction model; performing EOF reconstruction on an objective model prediction result; forming an integrated prediction result; and correction is carried out to realize climate anomaly space field prediction in different prediction time periods.

Embodiments provide drought projection for real-world geographic areas. One such embodiment identifies at least one climate model associated with a real-world geographic area. An evaluation is performed to determine whether the identified at least one climate model is valid for direct drought projection for the real-world geographic area. Based on a result of the evaluating, a drought projection technique is selected and the drought projection technique selected is employed to generate at least one drought projection for the real-world geographic area. The at least one drought projection includes an indication of projected drought frequency, projected drought duration, and projected drought intensity.

The invention discloses a comprehensive meteorological early warning system based on multi-source data fusion and artificial intelligence, and the system comprises a climate data collection module which is used for obtaining climate data; the climate data analysis module is used for receiving the climate data, calculating the climate data through an artificial intelligence algorithm and combining with a pre-constructed climate model to carry out climate prediction to obtain a prediction result; and the early warning module receives the prediction result and generates early warning information according to the prediction result. The historical climate data, the satellite remote sensing climate data and the real-time weather monitoring data are integrated, the climate data analysis module analyzes and processes the data to obtain the prediction result, and the early warning module generates the early warning information according to the prediction result, so that the prediction accuracy is ensured, the intelligence and the dynamics of the early warning mechanism are realized, and the early warning efficiency is improved. Powerful support can be provided for the development of strategic emerging industries.

The invention relates to a climate mode dynamic downscaling method introducing a dynamic SpectraNudging coefficient. The climate mode dynamic downscaling method comprises the following steps: (1) calculating a difference value between a background field variable and a mode simulation variable; (2) constructing a SpectraNING coefficient which dynamically changes along with time and space: fitting a cubic function which changes along with a vertical height layer of a climate mode, performing Z-score standardization and proper constraint, and traversing each horizontal grid point and different time of the mode to obtain the SpectraNING coefficient which dynamically changes along with time and three-dimensional space; (3) determining the influence height of the SpectraNudging coefficient; (4) calculating the time tendency of the climate mode simulation variable; and (5) repeatedly executing the steps (1)-(4), and continuously integrating the mode until the preset Nudging time is ended. The method can meet the temporal and spatial variation characteristics, so that the simulation description of the wind field is more detailed.

The invention discloses a U-Net-based multi-mode wind energy resource monthly scale prediction correction method and system, belongs to the technical field of wind energy resource evaluation and climate numerical value prediction crossing, and is used for correcting monthly scale wind speed output by a climate mode. According to the method, a climate state wind speed field is constructed by utilizing ERA5 reanalysis, and a monthly-scale 10m wind speed anomaly is calculated to serve as a correction reference; monthly-scale historical return data of a plurality of dynamic climate modes are obtained, 10m wind speed and multilayer meteorological elements are extracted, unified interpolation and standardization are carried out, and a sample set is formed; a U-Net correction model with a coding and decoding structure is constructed based on a sample set, feature combination and hyper-parameters are optimized through cross validation, and nonlinear mapping from a multi-mode forecast field to an ERA5 distance flat field is learned. And correcting future monthly scale forecast by using the optimal model, generating a wind speed product of which space structure and amplitude distribution are closer to observation, and providing high-credibility wind energy climate information for wind power planning, power generation planning and power grid dispatching.

Embodiments relate to computer systems and methods for computer models and scenario generation. The system involves generating integrated climate risk data using a Climate Risk Classification Standard hierarchy that maps climate data and multiple risk factors to geographic space and time. A computer model involves risk factors modeled as graphs of nodes, each node corresponding to a risk factor and connected by edges or links. The nodes of the graph create scenario paths for the model. The system automatically generates multifactor scenario sets using the scenario paths for the climate model to compute the likelihood of different scenario paths for the computer model. The scenario sets include transition scenarios.

The invention discloses a climate change prediction system based on a deep learning model, and the system comprises a data integration and processing module which is used for collecting, integrating and preprocessing multi-source climate data from a satellite, a ground observation station, climate model simulation and social media; the deep learning model construction and optimization module is used for constructing a multi-layer and multi-mode climate change prediction model based on a deep neural network architecture comprising a convolutional neural network, a recurrent neural network, a variant long-short term memory network of the recurrent neural network, a gating circulation unit and an attention mechanism; the model is optimized through transfer learning, hyper-parameter tuning and integrated learning methods; and the spatio-temporal feature extraction and fusion module efficiently extracts spatio-temporal features of climate data by using a deep learning model, and enhances the prediction ability of the model through a multi-source data fusion strategy including feature-level fusion and decision-level fusion.

The invention provides a building climate toughness evaluation method based on a climate model. According to the method, on the basis of building physical parameters, multi-source climate data and typical climate scene simulation results are fused, a multi-dimensional climate toughness index system including energy consumption response, comfort maintaining capacity and extreme event bearing capacity is constructed, and quantitative calculation and grading of the overall climate toughness of the building are achieved through a comprehensive evaluation model. Scientific basis and digital support are provided for the city to adapt to climate change, and the method has remarkable engineering popularization value and application prospect.

The invention discloses a regional extreme drought and flood event prediction system based on a climate model, and relates to the technical field of climate model prediction and disaster early warning, and the system comprises a data input module which is used for receiving regional multivariable prediction data processed by downscaling of the climate model, and the multivariable prediction data comprises rainfall, soil humidity and evapotranspiration. According to the regional extreme drought and flood event prediction system based on the climate model, equation strong constraint and variational optimization are carried out on key variables such as rainfall, soil humidity and evapotranspiration through the physical constraint dynamic coupler, the physical law of water circulation mass conservation and surface energy balance is forcibly met, the distortion phenomenon is eliminated, and the prediction accuracy is improved. According to the method, physical correction residual errors are converted into weight factors, Copula function parameters are dynamically adjusted, it is ensured that joint probability distribution of extreme drought and flood events strictly follows a physical mechanism, the path of misinformation physical impossible events is blocked from the source, and the problem of prediction distortion caused by multivariate physical inconsistency in downscaling output of a climate model is solved.

The invention relates to the technical field of meteorological data intelligent processing, and discloses a dynamic fusion data space downscaling method, system and device and a medium, and the method comprises the steps: obtaining first topographic data and first meteorological data, and improving the first meteorological data into second meteorological data through a self-adaptive dynamic upsampling operation; respectively extracting features of the first topographic data and the second meteorological data; fusing the extracted features based on a dynamic gating network, and performing a first enhancement operation to obtain optimized features; training a first learning model by adopting a mixed loss function; and based on the trained first learning model, performing convolution dimension reduction on the optimization features, and outputting a downscaled temperature field through a point-by-point convolution layer. The reconstruction precision of the method in a complex terrain area is obviously superior to that of a traditional interpolation method, high-resolution meteorological element products with continuous time and space can be generated, and high-precision data support is provided for refined weather forecast, climate model coupling and extreme weather event early warning.

The invention discloses an FY4A AGRI surface temperature-based mountain air temperature direct reduction rate spatial-temporal distribution estimation method, which comprises the following steps of: acquiring FY4A AGRI thermal infrared remote sensing surface temperature and influence factor data thereof, and preprocessing to obtain data under a 4km resolution ratio; performing spatial dimension reconstruction on the surface temperature missing data by adopting a random forest regression algorithm, and performing time dimension smoothing on a spatial dimension reconstruction result by adopting an S-G filter to obtain a surface temperature space-time dimension reconstruction result; constructing a surface temperature downscaling model by using a random forest regression algorithm, and reducing the resolution ratio of the surface temperature from 4km to 1km; the influence of the altitude on the surface temperature is separated, and the standardized surface temperature is calculated; and establishing a linear regression relationship between the standardized surface temperature and the altitude by adopting a sliding window method, and calculating a regression coefficient, so as to obtain an air temperature direct reduction rate. And necessary data and technical support are provided for related research fields such as weather forecast and climate model application.

The invention relates to the technical field of meteorological information processing, and particularly discloses a typical meteorological year data construction method for building energy consumption assessment, and the method comprises the steps: data collection, typical meteorological month selection, typical meteorological year data construction, and building energy consumption assessment. According to the scheme, the typical meteorological month is determined by combining the multi-dimensional day-by-day meteorological elements, the weights are given to the different meteorological elements, and the weighted average value is calculated, so that the dependence on a single element is avoided, the change trend of the meteorological elements in time is more comprehensively reflected through the combination of CDF and FS statistical values, the difference between long-term and short-term climate modes is considered, and the calculation accuracy is improved. A visual quantitative index is provided; according to the method, climate zones are divided based on building thermal load demands, missing hourly solar radiation data are complemented by using a mathematical model, so that a calculation result is more complete, the problem of data mutation caused by monthly boundaries is reduced by using smoothing processing, and the continuity and accuracy of data are improved.

The invention discloses a method for inverting a cloud condensation nucleus number concentration profile based on a laser radar and a neural network. The method comprises the following steps: collecting multi-wavelength laser radar data including a backscattering coefficient vertical profile, an extinction coefficient vertical profile, a laser radar ratio and a depolarization ratio; the method comprises the following steps: acquiring an atmospheric temperature profile and a water vapor mixing ratio profile which are in time-space synchronization with multi-wavelength laser radar data, performing grid alignment on a vertical height, and constructing a multi-channel input feature set for each preset target supersaturation ratio; and constructing a convolutional neural network profile inversion model, taking the multi-channel input feature set as input and the cloud condensation nucleus number concentration profile data as output, training the model, and obtaining the trained convolutional neural network profile inversion model. According to the method, the limitation of continuous, high-precision and high-temporal-spatial-resolution detection of CCN number concentration vertical distribution is overcome, and key data support is provided for the fields of weather forecast, climate models, artificial influence weather and the like.

The invention provides a cloud atmosphere high-fidelity vector radiation simulation method based on physical constraints. The method comprises the following steps: A1, generating cloud layer micro-physical parameters by a WRF model; a2, integrating simulation parameters generated in the step A1, and inputting the simulation parameters into an MYSTIC model for three-dimensional radiation transfer simulation; a3, constructing a PC-GAN (U-Net + + + differentiable radiation layer + double-branch discriminator) physical constraint generative adversarial correction network, and defining adversarial loss, physical loss and KL divergence loss; a4, simulation data are input into PC-GAN, CALIPSO, PLODER3 and other actual measurement data are compared with the simulation data to calculate loss, physical loss is fed back, and cloud micro physical parameters are iteratively optimized until RMSElt is radiated; 5%; a5, backfilling the cloud layer micro-physical parameters corrected in the step A4, carrying out MYSTIC model radiation transfer simulation again, and carrying out satellite / foundation verification; and A6, outputting high-fidelity radiation simulation data. The method can be widely applied to meteorological satellite data verification, climate model development, cloud characteristic inversion and remote sensor design, and has the technical advantages of high precision and high efficiency.

The application belongs to the field of sub-seasonal climate prediction, and provides a sub-seasonal climate prediction method and system based on the combination of dynamic model downscaling and machine learning downscaling, which comprises the following steps: S1, generating initial field and model boundary field information required for dynamic downscaling based on global climate model output data; S2, driving regional climate model to perform dynamic downscaling by using the initial field and model boundary field information; S3, generating input field required for machine learning downscaling based on the output circulation field information of the regional climate model; S4, performing machine learning downscaling correction optimization on the output circulation field of dynamic downscaling based on a convolution model; S5, performing machine learning super-resolution based on the output data of machine learning downscaling correction; and S6, generating sub-seasonal prediction information of double downscaling of dynamic model and machine learning. The application utilizes the complementary advantages of dynamic downscaling and deep learning downscaling, improves the sub-seasonal prediction skill, and copes with new challenges brought by climate change.

The application belongs to the technical field of data analysis and processing, and specifically discloses a typhoon track trend change confirmation method, system, device and medium. The method obtains the sea surface pressure disturbance field, the sea surface wind field and the diagnostic field in the sea basin scale. Based on the multi-field joint determination, the typhoon track and the influence area in the climate model data are determined. Further iteration is performed to determine the typhoon track and the influence area data set. Based on the modal decomposition and by using the influence area data set and the TDF method, the typhoon trend change is determined. The method can comprehensively utilize the forecast model data wind pressure field data to jointly research and judge, effectively determine the typhoon center position and the typhoon track, and objectively and reliably determine the typhoon influence range by the sea surface pressure disturbance field, so that the influence of the artificial factor introduced by the artificial given area weight influence factor in the TDF method can be reduced. The typhoon change trend is determined by using the principal component analysis method on the objective typhoon influence range, and the operability is high.

The present invention discloses a regional sub-seasonal wind energy resource prediction method, electronic device, storage medium and computer program product based on a dynamic climate model, aiming to improve the accuracy of sub-seasonal prediction of wind energy resources, especially to reduce the error of the dynamic climate model in the prediction of wind speed at the hub height of the wind turbine. The method analyzes the deviation characteristics of the model output and uses quantile regression technology to correct the error to achieve accurate prediction of wind speed within the next 1 to 45 days. When the present invention is implemented, the prediction ability of the model for the wind speed in the target area and the deviation characteristics for different levels of wind speed at the sub-seasonal scale are first analyzed. Secondly, according to the deviation characteristics of the model for different levels of wind speed, a correction scheme that can effectively improve the model deviation is selected, thereby improving the prediction ability of regional sub-seasonal scale wind energy resources and improving the prediction accuracy of the future evolution of regional wind energy resources.

The present application belongs to the technical field of aerosol measurement, and particularly relates to a method and system for measuring the surface tension of cloud droplets in the process of aerosol activation into clouds. The method comprises: building a system for measuring the surface tension of cloud droplets, including a cloud condensation nucleus counter, electrode plates, a high-voltage power supply, a condensation nucleus particle counter, and a differential mobility analyzer; applying a high-voltage electric field to cause positive and negative ions in the droplets to split under the traction of the electric field; aerosol samples are converted into droplets by absorbing water vapor in a cloud chamber, and the droplets split under the traction of the high-voltage electric field; the number concentration of the droplets in the outflow sample is measured by the condensation nucleus particle counter; the critical voltage at which the droplets split is obtained by fitting the graph of the number concentration of the droplets versus the voltage intensity; and the surface tension of the droplets is calculated according to the quantitative relationship between the critical voltage and the surface tension of the droplets. The present application can be used in atmospheric chemical transport simulation and climate models to evaluate the influence of the surface tension of cloud droplets on the number concentration and particle size of the cloud droplets.

The application discloses a kind of marine observation data gridding multi-level parallel acceleration calculation method and system.The application is based on the coarse-grained parallel implementation of climate model between MPI parallel architecture and parallel computing acceleration of ocean depth data, mainly includes the IO parallel reading of data, the calculation and distribution of climate model data, the synchronization of gridding data etc.;Based on the fine-grained parallel computing of OpenMP parallel architecture, the parallel computing of algorithm level between CPU core in node is realized, mainly includes the data gridding of each grid point in grid;For the gridding calculation of data in grid point, the distribution of data structure in memory is adjusted, and the specific calculation operation is accelerated using SIMD.The application can flexibly configure parallel scale according to specific hardware resources, realize the maximization of computing power utilization, and can provide computing power support for the construction of long-time sequence complete and reliable ocean grid data.

This invention discloses a generative deep learning-based downscaling correction method for climate model predictions. This method uses an unsupervised generative adversarial network to identify biases in global climate model hydrological and meteorological variable predictions. Using adversarial training of a generator and a discriminator, the adversarial loss is continuously reduced, enabling the trained generative deep learning network to handle complex nonlinear relationships, achieving a comprehensive optimization of downscaling and bias correction. This method can be used to correct the average bias of ensemble predictions of multi-model variables (precipitation, temperature, etc.) and improve their resolution. Inventive examples demonstrate that this method can effectively downscale multi-model ensemble predictions and achieves superior downscaling performance compared to commonly used quantile mapping and convolutional neural networks.

This invention discloses an inversion method for complex spatially distributed eddy viscosity coefficients in marine internal tide models, belonging to the field of marine numerical simulation technology. Key improvements of this invention include: the introduction of "independent point interpolation parameterization," a decisive solution to the computational efficiency bottleneck, transforming the problem from an "extremely difficult to compute" high-dimensional state to a "computable" low-to-medium-dimensional state, providing a feasible basis for subsequent operations; and the adoption of a "spectral projection gradient algorithm" as the optimization engine, further improving efficiency on the basis of dimensionality reduction, ensuring that the optimal solution is approximated with as few iterations as possible in each iteration step, thereby translating the advantages of dimensionality reduction into significant savings in actual computation time. This invention achieves a synergistic leap in accuracy, efficiency, and applicability, not only providing a powerful new tool for internal tide research but also potentially promoting the development of high-precision marine environmental forecasting and climate models, possessing significant scientific value and application potential.

The application discloses a satellite remote sensing data collection and preprocessing method and system, and relates to the technical field of satellite navigation. The method specifically comprises the following steps: collecting remote sensing images of a semiarid region through a satellite, analyzing pixel radiation and brightness gradient to identify bare land water features, constructing a leading atlas, extracting a boundary and registering to generate a standard structure, combining a climate model to correct to generate a reflectivity layer, transferring instructions when parameters are insufficient, reconstructing a queue by using a shortest algorithm to complete radiation calibration, atmospheric correction and geometric registration, and forming a preprocessing layer set. In the application, multi-band division is combined with radiation gradient calculation to enhance the separation degree of ground objects, boundary feature extraction and geometric registration are used to eliminate imaging distortion, a dynamic radiation model is fused with climate parameters to correct reflectivity, a double-threshold value is monitored to optimize a transmission strategy and avoid interruption, a priority algorithm is used to reconstruct a queue to reduce data delay, multi-level quality control is used to cooperatively process to form standard data, and the temporal and spatial reliability and application value are improved.

The present invention relates to the field of climate forecasting technology, and in particular to a multi-model ensemble sea temperature three-step sub-seasonal climate forecasting method and system. The method comprises the following steps: obtaining sea surface temperature data and sea ice coverage data, and pre-processing the sea surface temperature data and the sea ice coverage data to obtain standard input data; constructing an input field for an atmospheric circulation model based on the standard input data; inputting the input field into the atmospheric circulation model, and obtaining atmospheric circulation information and surface information based on the atmospheric circulation model; and performing dynamic downscaling based on a regional climate model according to the atmospheric circulation information and the surface information to obtain high-precision sub-seasonal climate forecast results. The present invention effectively solves the problem of insufficient accuracy of existing forecasting technologies, improves the accuracy and stability of sub-seasonal climate forecasts, provides new ideas for sub-seasonal climate forecasts, improves sub-seasonal climate forecasting techniques, and enhances the accuracy of forecast products.