Subseasonal to Decadal Prediction: Filling the Weather–Climate Gap

Merryfield, William J.; Baehr, Johanna; Batté, Lauriane; Becker, Emily; Butler, Amy H.; Coelho, Caio A. S.; Danabasoglu, Gökhan; Dirmeyer, Paul A.; Doblas‐Reyes, Francisco J.; Domeisen, Daniela I.V.; Ferranti, Laura; Ilynia, Tatiana; Kumar, Arun; Müller, Wolfgang A.; Rixen, M.; Robertson, Andrew W.; Smith, Doug; Takaya, Yuhei; Tuma, Matthias; White, Christopher J.; Alvarez, Mariano Sebastián; Ardilouze, Constantin; Attard, Hannah E.; Baggett, Cory F.; Balmaseda, Magdalena; Beraki, Asmerom F; Bhattacharjee, Partha; Bilbao, Roberto; Andrade, Felipe M. de; DeFlorio, Michael J.; Díaz, Leandro B.; Ehsan, Muhammad Fahad; Fragkoulidis, Georgios; Grainger, Sam; Green, Benjamin W.; Hell, Momme C.; Infanti, Johnna M.; Isensee, Katharina; Kataoka, Takahito; Kirtman, Ben P.; Klingaman, Nicholas P.; Lee, June‐Yi; Mayer, Kirsten; McKay, Roseanna C.; Mecking, Jennifer; Miller, Douglas E.; Neddermann, Nele; Ng, Ching Ho Justin; Ossó, Albert; Pankatz, Klaus; Peatman, Simon C.; Pegion, Kathy; Perlwitz, Judith; Recalde-Coronel, G. Cristina; Reintges, Annika; Renkl, Christoph; Solaraju-Murali, Balakrishnan; Spring, Aaron; Stan, Cristiana; Sun, Yongqiang; Tozer, Carly R.; Vigaud, Nicolas; Woolnough, Steven J.; Yeager, Stephen

doi:10.1175/bams-d-19-0037.a

Cited by 8 publications

(7 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Considering their huge socioeconomic impacts, reliable forecasts of weather and climate variations on various timescales are urgently required. However, owing to a gap between medium-range weather and seasonal forecasts (e.g., Vitart et al 2012;Merryfield et al 2020a), many challenges remain to perform seamless prediction across time scales (e.g., Luo et al 2016;Merryfield et al 2020b). To fill this gap, sustained efforts have been made by both the research and application communities to improve the subseasonal prediction (e.g., Brunet et al 2012;Mariotti et al 2018).…”

Section: Introductionmentioning

confidence: 99%

Improved MJO prediction using a multi-member subseasonal to seasonal forecast system of NUIST (NUIST CFS 1.1)

Luo

et al. 2023

Preprint

View full text Add to dashboard Cite

The Madden-Julian Oscillation (MJO) provides an important source of global subseasonal-to-seasonal (S2S) predictability, while its prediction remains great challenges. Based on an atmosphere-ocean coupled model and the widely-used nudging method, suitable initialization and ensemble schemes are explored toward an improved MJO prediction. Results show that the ensemble strategy with perturbed atmospheric nudging coefficients facilitates adequate ensemble spread and hence improves the prediction skill. Finally, an 18-member ensemble subseasonal prediction system called NUIST CFS1.1 is developed. Skill evaluation indicates that the NUIST CFS1.1 can extend the MJO prediction to 24 days lead, which reaches the world-average level but is far from the estimated potential predictability (~45 days). The limited skill at longer lead times corresponds to forecast errors exhibiting slower propagation and weaker intensity, which are largely owing to the model’s shortcoming in representing MJO-related physical processes. Despite the absence of dry bias, obvious biases in horizontal moisture gradients remain. Moreover, the model underestimates the diabatic heating (mostly contributed by the latent heat release) of enhanced convection and fails to reproduce the suppressed convection within the MJO structure, collaboratively weakening the Kelvin/Rossby waves. This causes weaker horizontal winds and ultimately reduces the horizontal moisture advection on the two flanks of MJO convection. Furthermore, the underestimated Kelvin wave induces insufficient planetary boundary layer (PBL) convergence and thereby results in poor simulation of PBL premoistening ahead of MJO convection. However, the relative strength between the Kelvin and Rossby wave keeps stationary as the forecast time increases, and thus is not a determining factor for MJO prediction in the NUIST CFS1.1. The above biases limit the MJO prediction, prompting further efforts to improve the model physics.

show abstract

Section: Introductionmentioning

confidence: 99%

Improved MJO prediction using a multi-member subseasonal to seasonal forecast system of NUIST (NUIST CFS 1.1)

Luo

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Accurate seasonal forecasts could bring enormous societal benefits and lead to improved agricultural productivity, water resource management, and disaster response planning (He et al, 2021). There is a growing interest from governments, industries, and non-governmental organizations (NGOs) in seasonal forecasts that could support early actions to weather extremes (Lang et al, 2020;Merryfield et al, 2020). As a consequence, seasonal timescales, lying between the short range weather forecasts and long range climate projections, have been receiving increasing attention by the forecast community (Bauer et al, 2015;Merryfield et al, 2020).…”

Section: Introductionmentioning

confidence: 99%

“…There is a growing interest from governments, industries, and non-governmental organizations (NGOs) in seasonal forecasts that could support early actions to weather extremes (Lang et al, 2020;Merryfield et al, 2020). As a consequence, seasonal timescales, lying between the short range weather forecasts and long range climate projections, have been receiving increasing attention by the forecast community (Bauer et al, 2015;Merryfield et al, 2020). However, traditional weather forecast models were not originally designed for those timescales (Cohen et al, 2019).…”

Section: Introductionmentioning

confidence: 99%

“…These include ocean temperatures, persistent atmospheric pressure patterns, land conditions, and sea ice (Mariotti et al, 2020). Despite the progress in the representation of these variables in weather models (Merryfield et al, 2020), error propagation still limits predictive power at seasonal scales (Cohen et al, 2019). These factors suggest a need for a different approach that can leverage the plethora of reanalysis data and observations from satellite remote sensing.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Seasonal prediction of Horn of Africa long rains using machine learning: The pitfalls of preselecting correlated predictors

et al. 2022

View full text Add to dashboard Cite

The Horn of Africa is highly vulnerable to droughts and floods, and reliable long-term forecasting is a key part of building resilience. However, the prediction of the “long rains” season (March–May) is particularly challenging for dynamical climate prediction models. Meanwhile, the potential for machine learning to improve seasonal precipitation forecasts in the region has yet to be uncovered. Here, we implement and evaluate four data-driven models for prediction of long rains rainfall: ridge and lasso linear regressions, random forests and a single-layer neural network. Predictors are based on SSTs, zonal winds, land state, and climate indices, and the target variables are precipitation totals for each separate month (March, April, and May) in the Horn of Africa drylands, with separate predictions made for lead-times of 1–3 months. Results reveal a tendency for overfitting when predictors are preselected based on correlations to the target variable over the entire historical period, a frequent practice in machine learning-based seasonal forecasting. Using this conventional approach, the data-driven methods—and particularly the lasso and ridge regressions—often outperform dynamical seasonal hindcasts. However, when the selection of predictors is done independently of both the train and test data, by performing this predictor selection within the cross-validation loop, the performance of all four data-driven models is poorer than that of the dynamical hindcasts. These findings should not discourage future applications of machine learning for rainfall forecasting in the region. Yet, they should be seen as a note of caution to prevent optimistically biased results that are not indicative of the true power in operational forecast systems.

show abstract

“…However, many critical applications -including water allocation, wildfire management, and drought and flood mitigation -require subseasonal forecasts (Fig. 5.1) with lead times between these two extremes, generally from 2 to 6 weeks in advance [117,190].…”

Section: Discussionmentioning

confidence: 99%

Balancing exploration and exploitation: task-targeted exploration for scientific decision-making

Flaspohler¹

View full text Add to dashboard Cite

How do we collect observational data that reveal fundamental properties of scientific phenomena? This is a key challenge in modern scientific discovery. Scientific phenomena are complex—they have high-dimensional and continuous state, exhibit chaotic dynamics, and generate noisy sensor observations. Additionally, scientific experimentation often requires significant time, money, and human effort. In the face of these challenges, we propose to leverage autonomous decision-making to augment and accelerate human scientific discovery. Autonomous decision-making in scientific domains faces an important and classical challenge: balancing exploration and exploitation when making decisions under uncertainty. This thesis argues that efficient decision-making in real-world, scientific domains requires task-targeted exploration—exploration strategies that are tuned to a specific task. By quantifying the change in task performance due to exploratory actions, we enable decision-makers that can contend with highly uncertain real-world environments, performing exploration parsimoniously to improve task performance. The thesis presents three novel paradigms for task-targeted exploration that are motivated by and applied to real-world scientific problems. We first consider exploration in partially observable Markov decision processes (POMDPs) and present two novel planners that leverage task-driven information measures to balance exploration and exploitation. These planners drive robots in simulation and oceanographic field trials to robustly identify plume sources and track targets with stochastic dynamics. We next consider the exploration- exploitation trade-off in online learning paradigms, a robust alternative to POMDPs when the environment is adversarial or difficult to model. We present novel online learning algorithms that balance exploitative and exploratory plays optimally under real-world constraints, including delayed feedback, partial predictability, and short regret horizons. We use these algorithms to perform model selection for subseasonal temperature and precipitation forecasting, achieving state-of-the-art forecasting accuracy. The human scientific endeavor is poised to benefit from our emerging capacity to integrate observational data into the process of model development and validation. Realizing the full potential of these data requires autonomous decision-makers that can contend with the inherent uncertainty of real-world scientific domains. This thesis highlights the critical role that task-targeted exploration plays in efficient scientific decision-making and proposes three novel methods to achieve task-targeted exploration in real-world oceanographic and climate science applications.

show abstract

Subseasonal to Decadal Prediction: Filling the Weather–Climate Gap

Abstract: This essay stems from international conferences on subseasonal to seasonal and seasonal to decadal prediction jointly convened by WWRP and WCRP in September 2018 in Boulder, Colorado: (www .wcrp-climate.org/s2s-s2d-2018-home).

Cited by 8 publications

References 0 publications

Improved MJO prediction using a multi-member subseasonal to seasonal forecast system of NUIST (NUIST CFS 1.1)

Improved MJO prediction using a multi-member subseasonal to seasonal forecast system of NUIST (NUIST CFS 1.1)

Seasonal prediction of Horn of Africa long rains using machine learning: The pitfalls of preselecting correlated predictors

Balancing exploration and exploitation: task-targeted exploration for scientific decision-making

Contact Info

Product

Resources

About