An official website of the United States government
Abstract The Southern Ocean is a region of high surface nutrient content, reflecting an inefficient biological carbon pump. The variability, predictability, and causes of changes in these nutrient levels on interannual to decadal time scales remain unclear. We employ a deep learning approach, specifically a Temporal Convolution Attention Neural Network (TCANN), to conduct multi‐year forecasting of surface based on oceanic physical drivers. The TCANN successfully replicates testing data with a prediction skill extending to at least 4 years with the GFDL‐ESM4‐driven model and 1 year with the observation‐driven model. To benchmark the results, we compare the prediction skill of TCANN with a simple persistence model and two regression methods, a linear regression and a ridge regression. The TCANN model was able to predict variability with a higher skill than persistence and the two regression methods indicating that non‐linearities present in the system become too high to predict inter‐annual variability with traditional regression methods. To enhance the interpretability of the predictions, we explore three explainable AI techniques: occlusion analysis, integrated gradients, and Gradient Shap. The outcomes suggest a crucial role played by salinity processes and buoyancy/potential density fluxes on the prediction of on annual time scales. The deep learning tools' ability to provide skillful forecasts well into the future presents a promising avenue for gaining insights into how the Southern Ocean's surface nutrients respond to climate change based on physical quantities.
more »
« less
This content will become publicly available on April 11, 2026
DroughtSet: Understanding Drought Through Spatial-Temporal Learning
Drought is one of the most destructive and expensive natural disasters, severely impacting natural resources and risks by depleting water resources and diminishing agricultural yields. Under climate change, accurately predicting drought is critical for mitigating drought-induced risks. However, the intricate interplay among the physical and biological drivers that regulate droughts limits the predictability and understanding of drought, particularly at a subseasonal to seasonal (S2S) time scale. While deep learning has demonstrated the potential to address climate forecasting challenges, its application to drought prediction has received relatively less attention. In this work, we propose a new dataset, DroughtSet, which integrates relevant predictive features and three drought indices from multiple remote sensing and reanalysis datasets across the contiguous United States (CONUS). DroughtSet specifically provides the machine learning community with a new real-world dataset to benchmark drought prediction models and more generally, time-series forecasting methods. Furthermore, we propose a spatial-temporal model SPDrought to predict and interpret S2S droughts. Our model learns from the spatial and temporal information of physical and biological features to predict three types of droughts simultaneously. Multiple strategies are employed to quantify the importance of physical and biological features for drought prediction. Our results provide insights for researchers to better understand the predictability and sensitivity of drought to biological and physical conditions. We aim to contribute to the climate field by proposing a new tool to predict and understand the occurrence of droughts and provide the AI community with a new benchmark to study deep learning applications in climate science.
more »
« less
- PAR ID:
- 10617868
- Publisher / Repository:
- AAAI Conference on Artificial Intelligence
- Date Published:
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract The effect of machine learning and other enhancements on statistical–dynamical forecasts of soil moisture (0–10 and 0–100 cm) and a reference evapotranspiration fraction [evaporative stress index (ESI)] on subseasonal time scales (15–28 days) are explored. The predictors include the current and past land surface conditions and dynamical model hindcasts from the Subseasonal to Seasonal Prediction project (S2S). When the methods are enhanced with machine learning and other improvements, the increases in skill are almost exclusively coming from predictors drawn from observations of current and past land surface states. This suggests that operational S2S flash drought forecasts should focus on optimizing use of information on current conditions rather than on integrating dynamically based forecasts, given the current state of knowledge. Nonlinear machine learning methods lead to improved skill over linear methods for soil moisture but not for ESI. Improvements for both soil moisture and ESI are realized by increasing the sample size by including surrounding grid points in training and increasing the number of predictors. In addition, all the improvements in the soil moisture forecasts predominantly impact soil moistening rather than soil drying—i.e., prediction of conditions moving away from drought rather than into drought—especially when the initial soil state is drier than normal. The physical reasons for the nonlinear machine learning improvements are also explored. Significance StatementRapidly intensifying droughts pose extra challenges for predictability. Here, dynamical forecast model output is combined with nonlinear machine learning methods to improve forecasts of rapid changes in soil moisture and the evaporative stress index (ESI).more » « less
-
Abstract The Madden‐Julian oscillation (MJO) is the leading source of global subseasonal predictability; however, many dynamical forecasting systems struggle to predict MJO propagation through the Maritime Continent. Better understanding the biases in simulated physical processes associated with MJO propagation is the key to improve MJO prediction. In this study, MJO prediction skill, propagation processes, and mean state biases are evaluated in reforecasts from models participating in the Subseasonal Experiment (SubX) and Subseasonal to Seasonal (S2S) prediction projects. SubX and S2S reforecasts show MJO prediction skill out to 4.5 weeks based on the Real‐time Multivariate MJO index consistent with previous studies. However, a closer examination of these models' representation of MJO propagation through the Maritime Continent reveals that they fail to predict the MJO convection, associated circulations, and moisture advection processes beyond 10 days with most of models underestimating MJO amplitude. The biases in the MJO propagation can be partly associated with the following mean biases across the Indo‐Pacific: a drier low troposphere, excess surface precipitation, more frequent occurrence of light precipitation rates, and a transition to stronger precipitation rates at lower humidity than in observations. This indicates that deep convection occurs too frequently in models and is not sufficiently inhibited when tropospheric moisture is low, which is likely due to the representation of entrainment.more » « less
-
Abstract The prediction skill for precipitation anomalies in late spring and summer months—a significant component of extreme climate events—has remained stubbornly low for years. This paper presents a new idea that utilizes information on boreal spring land surface temperature/subsurface temperature (LST/SUBT) anomalies over the Tibetan Plateau (TP) to improve prediction of subsequent summer droughts/floods over several regions over the world, East Asia and North America in particular. The work was performed in the framework of the GEWEX/LS4P Phase I (LS4P-I) experiment, which focused on whether the TP LST/SUBT provides an additional source for subseasonal-to-seasonal (S2S) predictability. The summer 2003, when there were severe drought/flood over the southern/northern part of the Yangtze River basin, respectively, has been selected as the focus case. With the newly developed LST/SUBT initialization method, the observed surface temperature anomaly over the TP has been partially produced by the LS4P-I model ensemble mean, and 8 hotspot regions in the world were identified where June precipitation is significantly associated with anomalies of May TP land temperature. Consideration of the TP LST/SUBT effect has produced about 25–50% of observed precipitation anomalies in most hotspot regions. The multiple models have shown more consistency in the hotspot regions along the Tibetan Plateau-Rocky Mountain Circumglobal (TRC) wave train. The mechanisms for the LST/SUBT effect on the 2003 drought over the southern part of the Yangtze River Basin are discussed. For comparison, the global SST effect has also been tested and 6 regions with significant SST effects were identified in the 2003 case, explaining about 25–50% of precipitation anomalies over most of these regions. This study suggests that the TP LST/SUBT effect is a first-order source of S2S precipitation predictability, and hence it is comparable to that of the SST effect. With the completion of the LS4P-I, the LS4P-II has been launched and the LS4P-II protocol is briefly presented.more » « less
-
Drought is a prominent feature of Hawaiʻi’s climate. However, it has been over 30 years since the last comprehensive meteorological drought analysis, and recent drying trends have emphasized the need to better understand drought dynamics and multi-sector effects in Hawaiʻi. Here, we provide a comprehensive synthesis of past drought effects in Hawaiʻi that we integrate with geospatial analysis of drought characteristics using a newly developed 100-year (1920–2019) gridded Standardized Precipitation Index (SPI) dataset. The synthesis examines past droughts classified into five categories: Meteorological, agricultural, hydrological, ecological, and socioeconomic drought. Results show that drought duration and magnitude have increased significantly, consistent with trends found in other Pacific Islands. We found that most droughts were associated with El Niño events, and the two worst droughts of the past century were multi-year events occurring in 1998–2002 and 2007–2014. The former event was most severe on the islands of O’ahu and Kaua’i while the latter event was most severe on Hawaiʻi Island. Within islands, we found different spatial patterns depending on leeward versus windward contrasts. Droughts have resulted in over $80 million in agricultural relief since 1996 and have increased wildfire risk, especially during El Niño years. In addition to providing the historical context needed to better understand future drought projections and to develop effective policies and management strategies to protect natural, cultural, hydrological, and agricultural resources, this work provides a framework for conducting drought analyses in other tropical island systems, especially those with a complex topography and strong climatic gradients.more » « less
