[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-126166-en":3,"doc-seo-126166-105":30,"detail-sidebar-cat-0-en-105":96},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":11,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},126166,3985741905716,"Rowan","https://ap-avatar.wpscdn.com/davatar_994ba38a5ba835b3df7d355c54d3ed8d",8,"Research & Report","An interpretable machine learning model for seasonal precipitation forecasting","Seasonal climate forecasting supports proactive decisions to mitigate adverse conditions or exploit favorable ones. The study introduces TelNet, a sequence-to-sequence model for short-to-medium lead seasonal precipitation forecasting that predicts an empirical precipitation distribution for each grid point using past precipitation and climate indices. TelNet uses a simple encoder-decoder-head design enabling training with limited data. Deterministic and probabilistic performance are evaluated against state-of-the-art dynamical and deep learning models, with resampled train/validation/test splits to quantify uncertainty and enable instance- and lead-wise interpretation.","| A Nature Portfolio journal |  |  |\n| --- | --- | --- |\n| [https://doi.org/10.1038/s43247-025-02207-2](https://doi.org/10.1038/s43247-025-02207-2) |  |  |\n| An interpretable machine learning model for seasonal precipitation forecasting\u003Cbr> Check for updates |  |  |\n| Enzo Pinheiro  & Taha B. M. J. Ouarda |  |  |\n| Seasonal climate forecasting is important for societal welfare, as it supports decision-makers in taking proactive steps to mitigate risks from adverse climate conditions or to take advantage of favorable ones. Here, we introduce TelNet, a sequence-to-sequence machine learning model for short-tomedium lead seasonal precipitation forecasting. The model takes past seasonal precipitation values and climate indices to predict an empirical precipitation distribution for every grid point of the target region for the next six overlapping seasons. TelNet has a simple encoder-decoder-head architecture, allowing the model to be trained with a limited amount of data, as is often the case in climate forecasting. Its deterministic and probabilistic performance is thoroughly evaluated and compared with state-of-the-art dynamical and deep learning models in a prominent region for seasonal forecasting studies due to its high climate predictability. The training, validation, and test sets are resampled multiple times to estimate the uncertainty associated with a small dataset. The results show thatTelNet ranks amongthe most accurate and calibrated models across multiple initialization months and lead times, especially during the rainy season when the predictable signal is strongest. Moreover, the model allows instance-and lead-wise forecast interpretation through its variable selection |  |  |\n| weights.\u003Cbr>Recent developments in sequence-to-sequence (seq2seq) machine learning models led to increased machine learning-based weather prediction (MLWP) models. The Convolutional Long-Short Term Memory (ConvLSTM) network was one of the first seq2seq models employed for precipitation nowcasting1. The model successfully captured the spatiotemporal patterns of its training dataset, outperforming previous stateof-the-art nowcasting models. Nevertheless, Long-Short Term Memory (LSTM) networks are limited in modeling very long sequences2. The selfattention mechanism, part of the transformer model3, solved these issues and led to a new generation of MLWP. For instance, MetNet was developed based on axial self-attention mechanism4 for probabilistic precipitation forecasts up to 8 h lead time and at a 1 km spatial resolution5. ClimaX, a foundation model for weather and climate modeling, was built upon vision transformer6 and able to cover a great range of tasks such as nowcasting, weather and subseasonal-to-seasonal (S2S) forecasting, and climate projections7. GraphCast8 was recently introduced for deterministic medium-range weather forecasts. It was based on Graph Neural Networks (GNN)9, 10, a network suitable for learning complex physics, such as weather dynamics. GenCast11, a diffusion model developed as an adaptation of GraphCast for probabilistic forecasting, also used transformer blocks in addition to GNNs. GraphCast and GenCast models outperformed the state-of-the-art European Centre for | Medium-Range Weather Forecasts (ECMWF) weather forecasting system in several cases.\u003Cbr>Numerical weather prediction (NWP) models, also known as dynamical models, map the current state of the atmosphere to future states by deterministically solving a set ofpartial differential equations that model the atmospheric dynamics. The rapid growth of uncertainty in the initial atmospheric condition and models’ imperfect representations12, 13 result in a limited forecasting horizon. Additionally, it is essential to track how the initial condition uncertainty evolves, which is done through ensemble forecasting. Ensemble forecasting uses the Monte Carlo method to approximate a stochastic dynamic forecast by repeatedly sampling from the initial condition probability di","cbCailN32AK6gQ6x","https://ap.wps.com/l/cbCailN32AK6gQ6x","pdf",3560162,1,14,"English","en",105,"# TelNet for seasonal precipitation forecasting\n## Model design and inputs\n## Deterministic and probabilistic evaluation\n## Uncertainty estimation with resampled datasets\n## Comparison with state-of-the-art dynamical and deep learning models\n## Interpretability for instance- and lead-wise forecasting","[{\"question\":\"What is TelNet designed to do?\",\"answer\":\"TelNet predicts an empirical precipitation distribution for each grid point in the target region for the next six overlapping seasons using past seasonal precipitation values and climate indices.\"},{\"question\":\"How does TelNet remain practical when climate datasets are limited?\",\"answer\":\"It uses a simple encoder-decoder-head architecture that can be trained with limited data, which is common in climate forecasting.\"},{\"question\":\"How is uncertainty handled in the model evaluation?\",\"answer\":\"Training, validation, and test sets are resampled multiple times to estimate the uncertainty associated with a small dataset.\"},{\"question\":\"What does the paper claim about interpretability?\",\"answer\":\"The model enables instance-wise and lead-wise forecast interpretation through variable selection weights.\"}]","An interpretable machine learning model for seasonal precipitation forecasting | PDF",1785903509,35,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":91,"head_meta":93,"extra_data":95,"updated_unix":28},"an-interpretable-machine-learning-model-for-seasonal-precipitation-forecasting","",{"@graph":36,"@context":90},[37,54,69],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/an-interpretable-machine-learning-model-for-seasonal-precipitation-forecasting/126166/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-23","2026-08-05",true,{"@type":66,"interactionType":67,"userInteractionCount":11},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82,86],{"name":73,"@type":74,"acceptedAnswer":75},"What is TelNet designed to do?","Question",{"text":76,"@type":77},"TelNet predicts an empirical precipitation distribution for each grid point in the target region for the next six overlapping seasons using past seasonal precipitation values and climate indices.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"How does TelNet remain practical when climate datasets are limited?",{"text":81,"@type":77},"It uses a simple encoder-decoder-head architecture that can be trained with limited data, which is common in climate forecasting.",{"name":83,"@type":74,"acceptedAnswer":84},"How is uncertainty handled in the model evaluation?",{"text":85,"@type":77},"Training, validation, and test sets are resampled multiple times to estimate the uncertainty associated with a small dataset.",{"name":87,"@type":74,"acceptedAnswer":88},"What does the paper claim about interpretability?",{"text":89,"@type":77},"The model enables instance-wise and lead-wise forecast interpretation through variable selection weights.","https://schema.org",{"og:url":52,"og:type":92,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":94,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":97},[98,102,106,110,115,120,125,128,133,136,140],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":99,"show_sort_weight":100,"slug":101},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":103,"show_sort_weight":104,"slug":105},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},"Exam",70,"exam",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},5,"Comic",60,"comic",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},6,"Technology",50,"technology",{"id":121,"doc_module":4,"doc_module_name":46,"category_name":122,"show_sort_weight":123,"slug":124},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":126,"slug":127},30,"research-report",{"id":129,"doc_module":4,"doc_module_name":46,"category_name":130,"show_sort_weight":131,"slug":132},9,"Religion & Spirituality",20,"religion-spirituality",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":134,"show_sort_weight":131,"slug":135},"World Cup","world-cup",{"id":137,"doc_module":4,"doc_module_name":46,"category_name":138,"show_sort_weight":137,"slug":139},10,"Lifestyle","lifestyle",{"id":141,"doc_module":4,"doc_module_name":46,"category_name":142,"show_sort_weight":111,"slug":143},19,"General","general"]