Andrew W. Robertson

Columbia University | CU · Earth Institute

We describe an innovative forecast presentation that aims to overcome obstacles to using seasonal climate forecasts for decision making, trace factors that influenced how seasonal forecast conventions have evolved, and describe a workshop process for training and supporting farmers in sub-Saharan Africa to use probabilistic seasonal forecasts. Main...

This paper assesses the skill of the Saudi-King Abdulaziz University coupled ocean–atmosphere Global Climate Model, namely Saudi-KAU CGCM, in forecasting the El Niño-Southern Oscillation (ENSO)-related sea surface temperature. The model performance is evaluated based on a reforecast of 38 years from 1982 to 2019, with 20 ensemble members of 12-mont...

This study evaluates the ability of state-of-the-art subseasonal to seasonal (S2S) forecasting systems to represent and predict the teleconnections of the Madden Julian Oscillations and their effects on weather in terms of midlatitude weather patterns and North Atlantic tropical cyclones. This evaluation of forecast systems applies novel diagnostic...

Predictability of Ethiopian Kiremt rainfall (June to September: JJAS) and forecast skill of the European Centre for Medium-Range Weather Forecasts (ECMWF) fifth-generation seasonal forecast system 5 (SEAS5) is explored during 1981–2019. The first empirical orthogonal function of observed rainfall explains 50.6% of the total variability and is chara...

The subseasonal-to-seasonal (S2S) predictive timescale, encompassing lead times ranging from 2 weeks to a season, is at the frontier of forecasting science. Forecasts on this timescale provide opportunities for enhanced application-focused capabilities to complement existing weather and climate services and products. There is, however, a ‘knowledge...

Subseasonal to seasonal (S2S) tropical rainfall predictability is assessed both from an analysis of the spatial scales of observed rainfall variability data, as well as from an S2S model reforecast skill. Observed spatial scales are quantified from gridded observed daily rainfall data, in terms of the size (area) of daily contiguous wet grid-points...

Skillful seasonal climate forecasts can support decision making in water resources management and agricultural planning. In arid and semi‐arid regions, tailoring reliable forecasts has the potential to improve water management by using key hydroclimate variables months in advance. This paper analyzes and compares the performance of two common appro...

This talk establishes a link between duration of local-scale wet events (defined from hourly rainfall records), area covered by rainfall (from IMERG data) and predictability (using a S2S ensemble from ECMWF) across India. Larger and longer wet events are more predictable than smaller and shorter ones. But shorter events still convey a significant a...

This talk compares some characteristics of daily rainfall (as mean daily intensity, wet patches area, spatial scales of rainfall anomalies, amplitude of variance conveyed by specific bandwidths) to skill and reproducibility at S2S timescales. It gives also details about S2S predictability across India.

Faced with the greatest public health crisis of our time, people must work together and learn from each other to overcome the complex challenges facing our communities, countries, and the world. Climate-related hazards are one of those challenges; they exacerbate already challenging public health conditions and impact not just people, but also the...

Recent research has highlighted the potential for improving predictive skill at the subseasonal timescale, which could be the basis for enhanced, actionable forecasts for climate services involving water and disaster management, health, energy and food security. Projects such as WMOʼs World Weather and World Climate Research Programmeʼs Subseasonal...

The Columbia University World Project “Adapting Agriculture to Climate Today, for Tomorrow” (ACToday), led by the International Research Institute for Climate and Society (IRI), was launched at the end of 2017 to create climate service solutions to help end hunger, achieve food security, improve nutrition, and promote sustainable agriculture in six...

Demands are growing rapidly in the operational prediction and applications communities for forecasts that fill the gap between daily weather forecasts and seasonal climate outlooks. Recent scientific advances have identified sources of predictability on this time range, and modeling advances are leading to better forecasts. However, much remains to...

Large-scale atmospheric circulation regime structures are used to diagnose subseasonal forecasts of wintertime geopotential height fields over the North American sector, from the NCEP CFSv2 model. Four large-scale daily circulation regimes derived from reanalysis 500hPa geopotential height data using K-means clustering are used as a low-dimensional...

Weather and climate variations on subseasonal to decadal time scales can have enormous social, economic, and environmental impacts, making skillful predictions on these time scales a valuable tool for decision-makers. As such, there is a growing interest in the scientific, operational, and applications communities in developing forecasts to improve...

There is high demand and a growing expectation for predictions of environmental conditions that go beyond 0–14-day weather forecasts with outlooks extending to one or more seasons and beyond. This is driven by the needs of the energy, water management, and agriculture sectors, to name a few. There is an increasing realization that, unlike weather f...

Calibrated probabilistic forecasts of weekly rainfall were developed for the state of Bihar innorthern India and issued in real time during the June–September 2018 monsoon period, up to 2 weeksin advance. The forecasts are based on subseasonal forecasts from the U.S. National Centers forEnvironmental Prediction CFSv2 model and were calibrated again...

Tropical rainfall is mostly convective and its subseasonal-to-seasonal (S2S) prediction remains challenging. We show that state-of-art model forecast skill 3-4 weeks ahead is systematically lower over land than ocean, which is matched by a similar land-ocean contrast in the spatial scales of observed biweekly rainfall anomalies. Regional difference...

This is the supplementary information of the main text in press in NPJ Atmospheric and Climate Science

This matlab dataset contains the onset and withdrawal dates of the boreal summer monsoon as explained in the paper. The onset dates are available in the matrix "Ons2" and the withdrawal dates are available in the matrix "End2". All dates refer to Jan 1st of the years 1901-2014 for 4954 grid-points defined by the LON/LAT vectors.

Successful climate services often involve the use of tailored regional climate forecasts at one or multiple timescales. The way those forecasts are implemented is not always straightforward, and depends on several different factors, like which variables, models and calibration methods to use, how to produce the ensemble and tailoring, or even how t...

The Subseasonal Experiment (SubX) is a multimodel subseasonal prediction experiment designed around operational requirements with the goal of improving subseasonal forecasts. Seven global models have produced 17 years of retrospective (re)forecasts and more than a year of weekly real-time forecasts. The reforecasts and forecasts are archived at the...

This paper reviews research done by the authors and their collaborators at IRI and beyond over the past decade on predictability and prediction of Indian summer monsoon rainfall (ISMR) on seasonal and sub-seasonal timescales. Empirical analyses of the daily ISMR characteristics at local scales pertinent to agriculture, based on IMD gridded data, re...

Considering lessons learned from experiences at seasonal timescale, this talk discusses some concrete S2S applications using both calibrated and uncalibrated forecasts from the S2S Database and the SubX project. First, we illustrate how a Python interface for IRI’s Climate Predictability Tool —PyCPT— can be employed to assess skill and calibrate su...

Intraseasonal timescales, more recently called sub-seasonal to seasonal (S2S), range from the deterministic limit of atmospheric predictability (about 10 days) up to a season (say, 100 days). These timescales occupy a window of overlap between low-frequency variability (LFV) intrinsic to the atmosphere and short-climatic timescales that also involv...

Common approaches to diagnose systematic errors involve the computation of metrics aimed at providing an overall summary of the performance of the model in reproducing the particular variables of interest in the study, normally tied to specific spatial and temporal scales. However, the evaluation of model performance is not always tied to the under...

Python interface for IRI’s Climate Predictability Tool (CPT), a widely used research and application Model Output Statistics/Prediction toolbox. Publicly available: GitHub. Automatically downloads required observations (TRMM, CPC Unified) and S2S model data from the IRI Data Library (S2S Database and SubX –ECMWF, CFSv2, GEFS, others are being in...

The aim of this research is to evaluate the temperature outputs of climate forecasting systems over Iran. The analysis is provided based on Atmosphere-Ocean Coupled General Circulation Models from North America Multi Model Ensemble (NMME). The skill of NMME individual models are evaluated in different initializations, of lead times (0-month, 1-mont...

The International Research Institute for Climate and Society Data Li- brary (IRIDL) is a powerful and freely accessible online data repository and analysis web-service that allows a user to view, analyze, and down- load hundreds of terabytes of climate-related data through a standard web browser in a computer or a smartphone. A wide variety of oper...

Common approaches to diagnose systematic errors involve the computation of metrics aimed at providing an overall summary of the performance of the model in reproducing the particular variables of interest in the study, normally tied to specific spatial and temporal scales. However, the evaluation of model performance is not always tied to the under...

Recent research has highlighted the potential for improving predictive skill at the sub-seasonal timescale, which could be the basis for enhanced, actionable forecasts for climate services involving water and disaster management, health, energy and food security. The WMO's World Weather and World Climate Research Programmes Subseasonal-to-Seasonal...

The sub-seasonal to seasonal prediction project (S2S) is a 5-year project, established in 2013 by the World Weather Research Program (WWRP) and the World Climate Research Program (WCRP). This paper briefl y describes the S2S project in the context of extended range prediction of extreme events. We provide evidence that S2S forecasts have the potent...

The daily characteristics of tropical rainfall, such as rainfall frequency, are important applications-relevant targets for sub-seasonal to seasonal (S2S) forecasts. Their potential predictability is assessed here based on observational estimates of spatial coherence of tropical rainfall anomalies estimated from the mean spatial autocorrelation in...

The aim of this research is to evaluate a statistical method for downscaling the precipitation output of a number of Coupled General Circulation Models issuing seasonal forecasts 9 month in advance. Canonical Correlation Analysis (CCA) is applied for post-processing precipitation from the North American Multi-model Ensemble (NMME) project. The anal...

An assessment is made of the ability of general circulation models in the CMIP5 ensemble to reproduce observed modes of low-frequency winter/spring precipitation variability in the region of the Upper Indus basin (UIB) in south-central Asia. This season accounts for about two thirds of annual precipitation totals in the UIB and is characterized by...

Abstract This paper provides a summary of the Workshop on Sub-Seasonal to Seasonal (S2S) Predictability of Extreme Weather and Climate, held at Columbia University, December 6–7, 2016. The 2-day workshop was attended by over 100 people and took stock of recent developments in Sub-seasonal to Seasonal predictability, S2S extreme weather phenomena, a...

The spatial coherence of interannual variations of sub-seasonal to seasonal anomalies in Indian summer monsoon rainfall is investigated at 0.25° spatial resolution using various metrics, including estimates of the number of degrees of freedom, the spatial scale of daily wet " patches ", as well as relationships between local and regional-scale rain...

Subseasonal forecast skill over the broadly defined North American (NAM), West African (WAM) and Asian (AM) summer monsoon regions is investigated using three Ensemble Prediction Systems (EPS) at sub-monthly lead times. Extended Logistic Regression (ELR) is used to produce probabilistic forecasts of weekly and week 3-4 averages of precipitation wit...

In this study we discuss probabilistic forecasts of Citarum River streamflow, which supplies 80 % of the water demands in Jakarta, Indonesia, based on general circulation model (GCM) output, for the September–November (SON) season. Retrospective forecasts of precipitation made over the period 1982–2010 with two coupled-ocean atmosphere GCMs, initia...

Streamflow forecasts at multiple time scales provide a new opportunity for reservoir management to address competing objectives. Market instruments such as forward contracts with specified reliability are considered as a tool that may help address the perceived risk associated with the use of such forecasts in lieu of traditional operation and allo...

Discrete-time hidden Markov models are a broadly useful class of latent-variable models with applications in areas such as speech recognition, bioinformatics, and climate data analysis. It is common in practice to introduce temporal non-homogeneity into such models by making the transition probabilities dependent on time-varying exogenous input var...

Any potential predictability will be of great value to cope with climate extremes, variability and change. Although hydro-energy, agricultural and water resource system planners require this information at seasonal to inter-annual time ranges, they have proven reluctant to incorporate climate forecasts into the decision making process. One reason l...

We propose a seamless diagnostic framework for coupled circulation models that can be used to catalogue their similarity to observations through timescales, and suggest model improvements.

Daily atmospheric circulation regimes can often be classified in robust clusters, or weather types (WTs), that describe available synoptic states of the region under study. Figuratively, these patterns can be seen to represent letters of an ‘alphabet’ that could be used to study the occurrence of impactful phenomena. For example, it has been shown...

This paper addresses the effect of interannual variability in jetstream orientation on weather systems over the North Atlantic basin (NAB). The observational analysis relies on 65 years of NCEP-NCAR reanalysis (1948–2012). The total daily kinetic energy of the geostrophic wind (GTKE) is taken as a measure of storm activity over the North Atlantic....

Potential and real predictive skill of the frequency of extreme rainfall in South East South America for the December-February season are evaluated in this paper, finding evidence indicating that mechanisms of climate variability at one timescale contribute to the predictability at another scale, i.e., taking into account the interference of differ...

The April 2011 floods in the Ohio River Basin and in the lower Mississippi River region were the latest of a set of major such flooding events recorded over the 20th century (defined in terms of a 10 year return maximum in stream flow). The questions of whether the recent 2011 event herald a return of more frequent flooding, and the degree of poten...

The physical mechanisms and predictability associated with extreme daily rainfall in South East South America (SESA) are investigated for the December-February season in a two-part study. Through a k-mean analysis, this first paper identifies a robust set of daily circulation regimes that are used to link the frequency of rainfall extreme events wi...

A Bayesian hidden Markov model (HMM) for climate downscaling of multisite daily precipitation is presented. A generalized linear model (GLM) component allows exogenous variables to directly influence the distributional characteristics of precipitation at each site over time, while the Markovian transitions between discrete states represent seasonal...

Daily rainfall occurrence and amount at 55 sta- tions over New Caledonia (NC, 20°S, 166°E) are exam- ined throughout the calendar year during 1980–2010 using a Hidden Markov Model (HMM). Daily rainfall variability is described in terms of six discrete rainfall states iden- tified by the HMM. Three states are interpreted as trade wind regimes associ...

Daily rainfall occurrence and amount at 55 stations over New Caledonia (NC, 20◦S, 166◦ E) are examined throughout the calendar year during 1980–2010 using a Hidden Markov Model (HMM). Daily rainfall variability is described in terms of six discrete rainfall states identified by the HMM. Three states are interpreted as trade wind regimes associated...

The prediction skill of precipitation at submonthly time scales during the boreal summer season is investigated based on hindcasts from three global ensemble prediction systems (EPSs). The results, analyzed for lead times up to 4 weeks, indicate encouraging correlation skill over some regions, particularly over the Maritime Continent and the equato...

A new World Weather Research Program/World Climate Research Program (WWRP/WCRP initiative on subseasonal to seasonal (S2S) prediction has recently been launched to foster collaboration and research in the weather and climate communities, with the goals of improving forecast skill and physical understanding, promoting forecast uptake by operational...

We present a Bayesian scheme for the downscaling of daily rainfall over a network of stations. Rainfall is modeled locally as a state-dependent mixture, with the states progressing in time as a first-order Markov process. The Markovian transition matrix, as well as the local state distributions, are dependent on exogenous covariates via generalized...