An evaluation of decadal probability forecasts from state-of-the-art climate models

While state-of-the-art models of Earth's climate system have improved tremendously over the last 20 years, nontrivial structural flaws still hinder their ability to forecast the decadal dynamics of the Earth system realistically. Contrasting the skill of these models not only with each other but also with empirical models can reveal the space and time scales on which simulation models exploit their physical basis effectively and quantify their ability to add information to operational forecasts. The skill of decadal probabilistic hindcasts for annual global-mean and regional-mean temperatures from the EU Ensemble-Based Predictions of Climate Changes and Their Impacts (ENSEMBLES) project is contrasted with several empirical models. Both the ENSEMBLES models and a “dynamic climatology” empirical model show probabilistic skill above that of a static climatology for global-mean temperature. The dynamic climatology model, however, often outperforms the ENSEMBLES models. The fact that empirical models display skill similar to that of today's state-of-the-art simulation models suggests that empirical forecasts can improve decadal forecasts for climate services, just as in weather, medium-range, and seasonal forecasting. It is suggested that the direct comparison of simulation models with empirical models becomes a regular component of large model forecast evaluations. Doing so would clarify the extent to which state-of-the-art simulation models provide information beyond that available from simpler empirical models and clarify current limitations in using simulation forecasting for decision support. Ultimately, the skill of simulation models based on physical principles is expected to surpass that of empirical models in a changing climate; their direct comparison provides information on progress toward that goal, which is not available in model–model intercomparisons.

Probabilistic skill in ensemble seasonal forecasts

Simulation models are widely employed to make probability forecasts of future conditions on seasonal to annual lead times. Added value in such forecasts is reflected in the information they add, either to purely empirical statistical models or to simpler simulation models. An evaluation of seasonal probability forecasts from the Development of a European Multimodel Ensemble system for seasonal to inTERannual prediction (DEMETER) and ENSEMBLES multi-model ensemble experiments is presented. Two particular regions are considered: Nino3.4 in the Pacific and the Main Development Region in the Atlantic; these regions were chosen before any spatial distribution of skill was examined. The ENSEMBLES models are found to have skill against the climatological distribution on seasonal time-scales. For models in ENSEMBLES that have a clearly defined predecessor model in DEMETER, the improvement from DEMETER to ENSEMBLES is discussed. Due to the long lead times of the forecasts and the evolution of observation technology, the forecast-outcome archive for seasonal forecast evaluation is small; arguably, evaluation data for seasonal forecasting will always be precious. Issues of information contamination from in-sample evaluation are discussed and impacts (both positive and negative) of variations in cross-validation protocol are demonstrated. Other difficulties due to the small forecast-outcome archive are identified. The claim that the multi-model ensemble provides a ‘better’ probability forecast than the best single model is examined and challenged. Significant forecast information beyond the climatological distribution is also demonstrated in a persistence probability forecast. The ENSEMBLES probability forecasts add significantly more information to empirical probability forecasts on seasonal time-scales than on decadal scales. Current operational forecasts might be enhanced by melding information from both simulation models and empirical models. Simulation models based on physical principles are sometimes expected, in principle, to outperform empirical models; direct comparison of their forecast skill provides information on progress toward that goal.

Towards a typology for constrained climate model forecasts

In recent years several methodologies have been developed to combine and interpret ensembles of climate models with the aim of quantifying uncertainties in climate projections. Constrained climate model forecasts have been generated by combining various choices of metrics used to weight individual ensemble members, with diverse approaches to sampling the ensemble. The forecasts obtained are often significantly different, even when based on the same model output. Therefore, a climate model forecast classification system can serve two roles: to provide a way for forecast producers to self-classify their forecasts; and to provide information on the methodological assumptions underlying the forecast generation and its uncertainty when forecasts are used for impacts studies. In this review we propose a possible classification system based on choices of metrics and sampling strategies. We illustrate the impact of some of the possible choices in the uncertainty quantification of large scale projections of temperature and precipitation changes, and briefly discuss possible connections between climate forecast uncertainty quantification and decision making approaches in the climate change context.

Origin and impact of initialisation shocks in coupled atmosphere-ocean forecasts

Current methods for initialising coupled atmosphere-ocean forecasts often rely on the use of separate atmosphere and ocean analyses, the combination of which can leave the coupled system imbalanced at the beginning of the forecast, potentially accelerating the development of errors. Using a series of experiments with the European Centre for Medium-range Weather Forecasts coupled system, the magnitude and extent of these so-called initialisation shocks is quantified, and their impact on forecast skill measured. It is found that forecasts initialised by separate ocean and atmospheric analyses do exhibit initialisation shocks in lower atmospheric temperature, when compared to forecasts initialised using a coupled data assimilation method. These shocks result in as much as a doubling of root-mean-square error on the first day of the forecast in some regions, and in increases that are sustained for the duration of the 10-day forecasts performed here. However, the impacts of this choice of initialisation on forecast skill, assessed using independent datasets, were found to be negligible, at least over the limited period studied. Larger initialisation shocks are found to follow a change in either the atmospheric or ocean model component between the analysis and forecast phases: changes in the ocean component can lead to sea surface temperature shocks of more than 0.5K in some equatorial regions during the first day of the forecast. Implications for the development of coupled forecast systems, particularly with respect to coupled data assimilation methods, are discussed.

A route to systematic error in forecasts of Rossby waves

Recent work has shown that both the amplitude of upper-level Rossby waves and the tropopause sharpness decrease with forecast lead time for several days in some operational weather forecast systems. In this contribution, the evolution of error growth in a case study of this forecast error type is diagnosed through analysis of operational forecasts and hindcast simulations. Potential vorticity (PV) on the 320-K isentropic surface is used to diagnose Rossby waves. The Rossby-wave forecast error in the operational ECMWF high-resolution forecast is shown to be associated with errors in the forecast of a warm conveyor belt (WCB) through trajectory analysis and an error metric for WCB outflows. The WCB forecast error is characterised by an overestimation of WCB amplitude, a location of the WCB outflow regions that is too far to the southeast, and a resulting underestimation of the magnitude of the negative PV anomaly in the outflow. Essentially the same forecast error development also occurred in all members of the ECMWF Ensemble Prediction System and the Met Office MOGREPS-15 suggesting that in this case model error made an important contribution to the development of forecast error in addition to initial condition error. Exploiting this forecast error robustness, a comparison was performed between the realised flow evolution, proxied by a sequence of short-range simulations, and a contemporaneous forecast. Both the proxy to the realised flow and the contemporaneous forecast a were produced with the Met Office Unified Model enhanced with tracers of diabatic processes modifying potential temperature and PV. Clear differences were found in the way potential temperature and PV are modified in the WCB between proxy and forecast. These results demonstrate that differences in potential temperature and PV modification in the WCB can be responsible for forecast errors in Rossby waves.

Assessing the accuracy and dispersion of real estate investment forecasts

Existing empirical evidence has frequently observed that professional forecasters are conservative and display herding behaviour. Whilst a large number of papers have considered equities as well as macroeconomic series, few have considered the accuracy of forecasts in alternative asset classes such as real estate. We consider the accuracy of forecasts for the UK commercial real estate market over the period 1999-2011. The results illustrate that forecasters display a tendency to under-estimate growth rates during strong market conditions and over-estimate when the market is performing poorly. This conservatism not only results in smoothed estimates but also implies that forecasters display herding behaviour. There is also a marked difference in the relative accuracy of capital and total returns versus rental figures. Whilst rental growth forecasts are relatively accurate, considerable inaccuracy is observed with respect to capital value and total returns.

Spatial evaluation of volcanic ash forecasts using satellite observations

The decision to close airspace in the event of a volcanic eruption is based on hazard maps of predicted ash extent. These are produced using output from volcanic ash transport and dispersion (VATD)models. In this paper an objectivemetric to evaluate the spatial accuracy of VATD simulations relative to satellite retrievals of volcanic ash is presented. The 5 metric is based on the fractions skill score (FSS). Thismeasure of skill provides more information than traditional point-bypoint metrics, such as success index and Pearson correlation coefficient, as it takes into the account spatial scale overwhich skill is being assessed. The FSS determines the scale overwhich a simulation has skill and can differentiate between a "near miss" and a forecast that is badly misplaced. The 10 idealised scenarios presented show that even simulations with considerable displacement errors have useful skill when evaluated over neighbourhood scales of 200–700km2. This method could be used to compare forecasts produced by different VATDs or using different model parameters, assess the impact of assimilating satellite retrieved ash data and evaluate VATD forecasts over a long time period.

Long-run restrictions and survey forecasts of output, consumption and investment

We consider the extent to which long-horizon survey forecasts of consumption, investment and output growth are consistent with theory-based steady-state values, and whether imposing these restrictions on long-horizon forecasts will enhance their accuracy. The restrictions we impose are consistent with a two-sector model in which the variables grow at different rates in steady state. The restrictions are imposed by exponential-tilting of simple auxiliary forecast densities. We show that imposing the consumption-output restriction yields modest improvements in the long-horizon output growth forecasts, and larger improvements in the forecasts of the cointegrating combination of consumption and output: the transformation of the data on which accuracy is assessed plays an important role.

Probabilistic wind power forecasts with an inverse power curve transformation and censored regression

Forecasting wind power is an important part of a successful integration of wind power into the power grid. Forecasts with lead times longer than 6 h are generally made by using statistical methods to post-process forecasts from numerical weather prediction systems. Two major problems that complicate this approach are the non-linear relationship between wind speed and power production and the limited range of power production between zero and nominal power of the turbine. In practice, these problems are often tackled by using non-linear non-parametric regression models. However, such an approach ignores valuable and readily available information: the power curve of the turbine's manufacturer. Much of the non-linearity can be directly accounted for by transforming the observed power production into wind speed via the inverse power curve so that simpler linear regression models can be used. Furthermore, the fact that the transformed power production has a limited range can be taken care of by employing censored regression models. In this study, we evaluate quantile forecasts from a range of methods: (i) using parametric and non-parametric models, (ii) with and without the proposed inverse power curve transformation and (iii) with and without censoring. The results show that with our inverse (power-to-wind) transformation, simpler linear regression models with censoring perform equally or better than non-linear models with or without the frequently used wind-to-power transformation.

An assessment of Indian monsoon seasonal forecasts and mechanisms underlying monsoon interannual variability in the Met Office GloSea5-GC2 system

We assess Indian summer monsoon seasonal forecasts in GloSea5-GC2, the Met Office fully coupled subseasonal to seasonal ensemble forecasting system. Using several metrics, GloSea5-GC2 shows similar skill to other state-of-the-art forecast systems. The prediction skill of the large-scale South Asian monsoon circulation is higher than that of Indian monsoon rainfall. Using multiple linear regression analysis we evaluate relationships between Indian monsoon rainfall and five possible drivers of monsoon interannual variability. Over the time period studied (1992-2011), the El Nino-Southern Oscillation (ENSO) and the Indian Ocean dipole (IOD) are the most important of these drivers in both observations and GloSea5-GC2. Our analysis indicates that ENSO and its teleconnection with the Indian rainfall are well represented in GloSea5-GC2. However, the relationship between the IOD and Indian rainfall anomalies is too weak in GloSea5-GC2, which may be limiting the prediction skill of the local monsoon circulation and Indian rainfall. We show that this weak relationship likely results from a coupled mean state bias that limits the impact of anomalous wind forcing on SST variability, resulting in erroneous IOD SST anomalies. Known difficulties in representing convective precipitation over India may also play a role. Since Indian rainfall responds weakly to the IOD, it responds more consistently to ENSO than in observations. Our assessment identifies specific coupled biases that are likely limiting GloSea5-GC2 prediction skill, providing targets for model improvement.

Seasonal forecasts of tropical cyclone activity in a high-atmospheric-resolution coupled prediction system

Seasonal forecast skill of the basinwide and regional tropical cyclone (TC) activity in an experimental coupled prediction system based on the ECMWF System 4 is assessed. As part of a collaboration between the Center for Ocean–Land–Atmosphere Studies (COLA) and the ECMWF called Project Minerva, the system is integrated at the atmospheric horizontal spectral resolutions of T319, T639, and T1279. Seven-month hindcasts starting from 1 May for the years 1980–2011 are produced at all three resolutions with at least 15 ensemble members. The Minerva system demonstrates statistically significant skill for retrospective forecasts of TC frequency and accumulated cyclone energy (ACE) in the North Atlantic (NA), eastern North Pacific (EP), and western North Pacific. While the highest scores overall are achieved in the North Pacific, the skill in the NA appears to be limited by an overly strong influence of the tropical Pacific variability. Higher model resolution improves skill scores for the ACE and, to a lesser extent, the TC frequency, even though the influence of large-scale climate variations on these TC activity measures is largely independent of resolution changes. The biggest gain occurs in transition from T319 to T639. Significant skill in regional TC forecasts is achieved over broad areas of the Northern Hemisphere. The highest-resolution hindcasts exhibit additional locations with skill in the NA and EP, including land-adjacent areas. The feasibility of regional intensity forecasts is assessed. In the presence of the coupled model biases, the benefits of high resolution for seasonal TC forecasting may be underestimated.

Exploiting existing ground-based remote sensing networks to improve high-resolution weather forecasts

A new generation of high-resolution (1 km) forecast models promises to revolutionize the prediction of hazardous weather such as windstorms, flash floods, and poor air quality. To realize this promise, a dense observing network, focusing on the lower few kilometers of the atmosphere, is required to verify these new forecast models with the ultimate goal of assimilating the data. At present there are insufficient systematic observations of the vertical profiles of water vapor, temperature, wind, and aerosols; a major constraint is the absence of funding to install new networks. A recent research program financed by the European Union, tasked with addressing this lack of observations, demonstrated that the assimilation of observations from an existing wind profiler network reduces forecast errors, provided that the individual instruments are strategically located and properly maintained. Additionally, it identified three further existing European networks of instruments that are currently underexploited, but with minimal expense they could deliver quality-controlled data to national weather services in near–real time, so the data could be assimilated into forecast models. Specifically, 1) several hundred automatic lidars and ceilometers can provide backscatter profiles associated with aerosol and cloud properties and structures with 30-m vertical resolution every minute; 2) more than 20 Doppler lidars, a fairly new technology, can measure vertical and horizontal winds in the lower atmosphere with a vertical resolution of 30 m every 5 min; and 3) about 30 microwave profilers can estimate profiles of temperature and humidity in the lower few kilometers every 10 min. Examples of potential benefits from these instruments are presented.

A prospectus for future geomorphological investigation of the Namib sand sea

The Namib Sand Sea in southern Africa offers an ideal location in which to consider general questions about the evolution of sand seas, about the fluxes of sand through contemporary dune fields and about the patterns of dune form that are created. This paper aims to provide a concise account of the approaches and techniques that are currently being used and will be used in the future to address these questions. The paper considers the techniques employed to investigate wind climate, the morphometry of the dunes, the internal structure of dune sediments, the age of the dunes and the potential to model dune development

Extended time weather forecasts contributes to agricultural productivity estimates

Weather conditions in critical periods of the vegetative crop development influence crop productivity, thus being a basic parameter for crop forecast. Reliable extended period weather forecasts may contribute to improve the estimation of agricultural productivity. The production of soybean plays an important role in the Brazilian economy, because this country is ranked among the largest producers of soybeans in the world. This culture can be significantly affected by water conditions, depending on the intensity of water deficit. This work explores the role of extended period weather forecasts for estimating soybean productivity in the southern part of Brazil, Passo Fundo, and Londrina (State of Rio Grande do Sul and Parana, respectively) in the 2005/2006 harvest. The goal was to investigate the possible contribution of precipitation forecasts as a substitute for the use of climatological data on crop forecasts. The results suggest that the use of meteorological forecasts generate more reliable productivity estimates during the growth period than those generated only through climatological information.

Evaluation Of Reservoir Model Integration With Deterministic And Probabilistic Streamflow Forecasts

Due to the increase in water demand and hydropower energy, it is getting more important to operate hydraulic structures in an efficient manner while sustaining multiple demands. Especially, companies, governmental agencies, consultant offices require effective, practical integrated tools and decision support frameworks to operate reservoirs, cascades of run-of-river plants and related elements such as canals by merging hydrological and reservoir simulation/optimization models with various numerical weather predictions, radar and satellite data. The model performance is highly related with the streamflow forecast, related uncertainty and its consideration in the decision making. While deterministic weather predictions and its corresponding streamflow forecasts directly restrict the manager to single deterministic trajectories, probabilistic forecasts can be a key solution by including uncertainty in flow forecast scenarios for dam operation. The objective of this study is to compare deterministic and probabilistic streamflow forecasts on an earlier developed basin/reservoir model for short term reservoir management. The study is applied to the Yuvacık Reservoir and its upstream basin which is the main water supply of Kocaeli City located in the northwestern part of Turkey. The reservoir represents a typical example by its limited capacity, downstream channel restrictions and high snowmelt potential. Mesoscale Model 5 and Ensemble Prediction System data are used as a main input and the flow forecasts are done for 2012 year using HEC-HMS. Hydrometeorological rule-based reservoir simulation model is accomplished with HEC-ResSim and integrated with forecasts. Since EPS based hydrological model produce a large number of equal probable scenarios, it will indicate how uncertainty spreads in the future. Thus, it will provide risk ranges in terms of spillway discharges and reservoir level for operator when it is compared with deterministic approach. The framework is fully data driven, applicable, useful to the profession and the knowledge can be transferred to other similar reservoir systems.