Previous Forecasts Research Articles

Convergent least-squares optimization methods for variational data assimilation

Data assimilation combines prior (or background) information with observations to estimate the initial state of a dynamical system over a given time-window. A common application is in numerical weather prediction where a previous forecast and atmospheric observations are used to obtain the initial conditions for a numerical weather forecast. In four-dimensional variational data assimilation (4D-Var), the problem is formulated as a nonlinear least-squares problem, usually solved using a variant of the classical Gauss-Newton (GN) method. However, we show that GN may not converge if poorly initialized. In particular, we show that this may occur when there is greater uncertainty in the background information compared to the observations, or when a long time-window is used in 4D-Var allowing more observations. The difficulties GN encounters may lead to inaccurate initial state conditions for subsequent forecasts. To overcome this, we apply two convergent GN variants (line search and regularization) to the long time-window 4D-Var problem and investigate the cases where they locate a more accurate estimate compared to GN within a given budget of computational time and cost. We show that these methods are able to improve the estimate of the initial state, which may lead to a more accurate forecast. Highlights Poor initialization of Gauss-Newton method may result in failure to converge. Safeguarded Gauss-Newton improves initial state estimate within limited time/cost. Results using twin experiments with long time-window and chaotic Lorenz models. Apply state of the art least-squares convergence theory to data assimilation. Improvements to initial state estimate may lead to a more accurate forecast.

Optimal constraints on Primordial non-Gaussianity with the eBOSS DR16 quasars in Fourier space

We present constraints on the amplitude of local Primordial Non-Gaussianities (PNG), f NL, using the quasar sample in the Sloan Digital Sky Survey IV extended Baryon Oscillation Spectroscopic Survey (eBOSS) Data Release 16 (DR16). We analyze the power spectrum monopole, testing for the presence of scale-dependent galaxy bias induced by local PNG. Our analysis makes use of optimal redshift weights that maximize the response of the quasar sample to the possible presence of non-zero PNG. We find -4 < f NL < 27 at 68% confidence level, which is among the strongest bounds with Large Scale Structure data. The optimal analysis reduces the error bar by ~10% compared to the standard one, but this improvement is lower than the one expected from previous forecasts. In addition, the larger volume of this dataset, when compared to previous releases of the eBOSS quasar catalog, does not always correspond to a reduction of the final uncertainty on local PNG. This could suggest the presence of still unknown systematic effects in the data.If the quasars have a lower response to local PNG, our optimal constraint becomes -23 < f NL < 21 at 68%, with an improvement of 30% over standardanalyses. We alsoshow how to use the optimal weights to put data-driven priors on the sample's response tolocal PNG.

Scaling up solar and wind electricity: empirical modelling and a disruptive scenario for their deployments in Aotearoa New Zealand

ABSTRACT Deployment of wind and solar electricity technologies is crucial for the energy transition, yet anticipated deployment rates differ widely often underestimating actual deployments. This issue is significant for countries like Aotearoa New Zealand, leading the way in major energy transitions. This study addresses two main issues: the historical underestimation of deployment rates and the need to synthesise various future scenarios to assess their potential impacts on the energy system. To address these problems, an empirical modelling approach based on logistic growth is adopted to create and analyse scenarios for the deployment of solar and wind technologies in New Zealand. We propose two scenarios: a reference growth scenario, which represents prevailing estimations for the deployment of solar and wind from previous forecasts; and a disruptive scenario, which anticipates a more rapid deployment pace. Both scenarios envision phasing out fossil fuels by 2030 but differ in the remaining electricity supply. In the disruptive scenario, the installed electricity generation broadens, in which solar and wind generation could represent 72% of the electricity generation capacity by 2050. Our findings emphasise the transformative potential of rapid wind and solar technologies deployment and highlight the importance of preparing for a disruptive future that challenges conventional expectations.

Signalling with Private Monitoring

Abstract We study dynamic signalling when the sender does not see the signals that her actions generate. The sender then uses her past play to forecast what a receiver believes, in turn forcing the receiver to forecast the previous forecast, and so forth. We identify a class of linear-quadratic-Gaussian games where this endogenous higher-order uncertainty can be handled. The sender’s second-order belief is key: it is a private state that she controls, and it creates a new channel for information transmission. We examine the role of higher-order uncertainty and this new signalling channel in applications to macroeconomics, reputation, and trading: inflationary biases under discretion can be larger; career-concerned agents may benefit from not knowing their reputations; and informed trades can carry more price impact. We also introduce an existence method for boundary value problems that can be used in other dynamic games.

Climatic disasters and distracted analysts

AbstractUsing a sample of 30,270 forecasts by 2,280 analysts under a stacked difference‐in‐differences framework involving 22 major climatic disasters in the United States, we examine the effect of climatic disasters on information production by security analysts, who play a crucial role in the financial market as information intermediaries. We find that earnings forecasts by analysts who experienced a major climatic disaster become less accurate than those by unaffected analysts within 3 months after the disaster, due to distracted attention. Disaster‐zone analysts are more likely to allocate their attention to firms of greater importance or salience, and they tend to reiterate their previous forecasts to maintain the quantity and timeliness of their forecasts. Overall, we document the real impact of cognitive bias on financial professionals' performance.

Повышение качества прогнозирования простейшими методами комбинирования отдельных прогнозов

Combining forecasts is considered the easiest way to improve the forecast quality compared to individual models. In this paper, we test the capabilities of the simplest methods of combination, such as simple averages and estimates based on the standard error of previous forecasts, to improve the performance of short-run forecasts of five resource price indicators (oil and metals). The basis of the work is the Gaidar Institute forecasts database, which provides the database of primary forecasts and allows you to calculate their combinations in real time. Based on the obtained results we conclude that even the simplest methods of combination are a way to improve the accuracy of forecasts. In addition, in the case of resource prices, one can even single out a group of methods (namely, combining with weights inversely proportional to the squared errors of individual forecasts) that provide the maximum gain in quality for the most periods.

Towards a full wCDM map-based analysis for weak lensing surveys

ABSTRACT The next generation of weak lensing surveys will measure the matter distribution of the local universe with unprecedented precision, allowing the resolution of non-Gaussian features of the convergence field. This encourages the use of higher-order mass-map statistics for cosmological parameter inference. We extend the forward-modelling based methodology introduced in a previous forecast paper to match these new requirements. We provide multiple forecasts for the $w$CDM parameter constraints that can be expected from stage 3 and 4 weak lensing surveys. We consider different survey setups, summary statistics and mass map filters including wavelets. We take into account the shear bias, photometric redshift uncertainties, and intrinsic alignment. The impact of baryons is investigated and the necessary scale cuts are applied. We compare the angular power spectrum analysis to peak and minima counts as well as Minkowski functionals of the mass maps. We find a preference for Starlet over Gaussian filters. Our results suggest that using a survey setup with 10 instead of 5 tomographic redshift bins is beneficial. Adding cross-tomographic information improves the constraints on cosmology and especially on galaxy intrinsic alignment for all statistics. In terms of constraining power, we find the angular power spectrum and the peak counts to be equally matched for stage 4 surveys, followed by minima counts and the Minkowski functionals. Combining different summary statistics significantly improves the constraints and compensates the stringent scale cuts. We identify the most ‘cost-effective’ combination to be the angular power spectrum, peak counts and Minkowski functionals following Starlet filtering.

Implementation of Hidden Infections in the SEIR Model Analysis of the Omicron Variant Spread in Bulgaria

In this paper we study the impact of the hidden cases of Covid-19 spread in Bulgaria by means of the SEIR model, in the period December 15, 2021 – until March 2022. In particular, the numerical experiments based on the hidden cases till mid-January, provide a forecast of the peak of the infectious curve about February 8, 2022, and reaching the maximum of about 90 000 cases. The forecast shows that end of the epidemics (in Bulgaria) would be in mid-May, 2022. This forecast is compared to the previous forecasts based on the same data excluding hidden cases; they have shown that the peak will be about February 28, reaching about 75 000 cases and will vanish until end of May. As seen in Fig. 3, our forecast of the end of the particular wave of epidemics shows just two weeks difference with what happened in real life – the wave ended in the first week of June. It is curious, that though the peak of the curve in thehidden cases approach has shifted to February 8, the cumulative number of cases in both approaches is nearly the same. More details on the methods used in the present paper and the non-hidden cases study, are provided as well.

Testing the Lee-Carter model on Hungarian mortality data

AbstractThis article tests the popular Lee-Carter model's performance for Hungarian mortality rate forecasting. Hungary passed through a mortality crisis which makes the task particularly difficult. Previous forecasts and model choices are validated, and updated forecasts are produced. We find that the behaviour of mortality rates is normalizing, and so the basic Lee-Carter model is becoming applicable.

Heterogeneous expectations, forecast accuracy and firms’ credit demand

Firm heterogeneity in formulating expectations constitutes a source of friction for the credit market. An expected improvement of credit conditions in the future may slacken current credit demand. Using a panel of European firms in the period 2010–2019, I find that better expectations on future (six-month) access to credit reduce the probability of a firm to apply for bank loans by 3 percent. Results also suggest that access to credit expectations are adaptive. Although firms fail 43 percent of their access to credit forecasts, they adjust their expectations learning from previous forecast errors.

RETRACTED: Forecasting Solar Cycle 25 Using an Optimized Long Short-term Memory Mode Based on F10.7 and Sunspot Area Data

In this paper, an optimized long short-term memory model is proposed to deal with the smoothed monthly F 10.7 and nonsmoothed monthly sunspot area (SSA) data, aiming to forecast the peak amplitude of both solar activities and the occurring time for Solar Cycle 25 (SC-25), as well as to obtain the maximum amplitude of sunspot number (SSN) and the reaching time according to the relationships between them. The “reforecast” process in the model uses the latest forecast results obtained from the previous forecast as the input for the next forecasting calculation. The forecasting errors between the forecast and observed peak amplitude of F 10.7 for SC-23 and SC-24 are 2.87% and 1.09%, respectively. The results of this evaluation indicator of SSA for SC-21 to SC-24 were 8.85%, 4.49%, 2.88%, and 4.57%, respectively, and the errors for the occurring time were all within 6 months. The forecast peak amplitude of F 10.7 and SSA for SC-25 is 156.3 and 2562.5 respectively, and the maximum values of SSN are calculated as 147.9 and 213 based on F 10.7 and SSA respectively, which implies that SC-25 will be stronger than SC-24, and that SC-25 will reach its peak at the beginning of 2025.

ИЗМЕНЕНИЕ ДОЛГОСРОЧНЫХ ПРОГНОЗОВ РАЗВИТИЯ МИРОВОГО РЫНКА НЕФТИ ПОД ВЛИЯНИЕМ НОВОГО ГЕОПОЛИТИЧЕСКОГО ФАКТОРА

Predicting the global oil market for the long term, both oil companies and analytical centers use a systematic approach, striving to take into account all the main factors affecting the sensitive oil market. Each new factor affects the market in a new way, setting new vectors of development or restrictions. The climate factor that had been in effect until recently, which prompted the global energy market to restructure taking into account the gradual reduction of CO2 emissions, was replaced by a pandemic factor that limited the development of demand for oil. The latest recent factor was the factor of political decisions related to the escalation of the conflict between Russia and Ukraine, which resulted in unprecedented sanctions pressure on the Russian economy. The forecasting entities represented by BP Oil Company, the International Energy Agency and OPEC have once again revised their forecasts for the development of the oil market. This article is devoted to the analysis of these forecasts, as well as comparison with previous forecasts made after the pandemic.

The air cargo market overview

Measured by the value of goods, about one-third of all international trade is moved by air. It stands for a big chunk of the transport market and global GDP, playing a crucial role in moving products of high value in relation to their weight, but also the backbone of overnight shipping and enabling e-commerce growth. This places air cargo as very dependent on overall economic deviations. As the world slowly exits the global pandemic state, each part of the aviation industry should be subject to analyses that confirm or contradict previous forecasts, thus helping to make correct business decisions by the relevant entities in the aviation industry. The following article is devoted to the analysis of the air cargo market. As a starting point, the article shows a general overview of the world's economy by pointing out the main variables that impact demand for air cargo and presenting forecasts on some of those. General air cargo market overview is the next subject. This part shows the latest trends connected with the general aviation market and the cargo part, outlining the general look. An overview of forecasts for the aviation market, coming from aircraft manufacturers, is the last of the analytical parts of this article, describing each entity's market outlook. Presented analyses were later used to determine trends most likely to show in coming years. The accurate description of those allowed for creating a coherent forecast of the air cargo market, with the calculation of actual cargo tonne-kilometers for oncoming years using a simple, multivariate forecasting method based on creating a historical data-driven model. The article concludes with a summary that provides an overview of the covered subjects.

Stress State of the Earth’s Crust, Seismicity, and Prospects for Long-Term Forecast of Strong Earthquakes in Uzbekistan

Abstract —The current stress state of the Earth’s crust on the territory of Uzbekistan is studied using algorithms from the cataclastic analysis of displacements (CAD) with the application of a combined catalog of earthquake source mechanisms. The axis of maximum compression is nearly horizontal and orthogonal to the strike of tectonic structures for the larger part of the area under study, and the angle of penetration of the axis of minimum compression varies greatly for different parts of the territory. The areal distribution of the Lode–Nadai coefficient indicates the predominance of a stress state close to pure shift for the area under study. Without differentiation of the seismically active bed by depth, almost the entire territory of Uzbekistan is characterized by a geodynamic stress state corresponding to horizontal compression. There are significant differences in the stress state parameters for different deep beds of the Earth’s crust. With account for the representative recording periods of seismic events of different energy levels, the earthquake recurrence parameters for the entire territory as a whole and large seismically active zones are determined. It has been established that the epicenters of strong earthquakes recorded since the historical time are grouped in compact areas with linear dimensions of 50–80 km. The tectonophysical interpretation of the selected areas is given within the framework of the CAD, in which it is noted that strong earthquakes occur mainly in areas of lower effective confining pressure and maximum tangent stresses. This is due to the fact that the values of friction forces at faults in such areas are low, which creates the most favorable conditions for large-scale destruction. According to the results of the reconstruction of the natural stresses of the CAD, such areas on the territory of Uzbekistan are identified within faults and flexural-fault zones in South Fergana and North Fergana, as well as in the Gazli region. The current seismological situation in the selected areas of long-term forecast is estimated using a set of prognostic parameters of the seismic regime, and a map of the areas of expected seismic activity is compiled. The analysis of previous forecast maps developed within the framework of the presented approach shows high value in their information.

Projection of primary and revision hip arthroplasty surgery in Denmark from 2020 to 2050.

The incidence of primary and revision total hip arthroplasty (THA) has increased over the last decades. Previous forecasts from different healthcare systems have predicted a continuous increase. We present a forecast of both primary and revision surgery from 2020 to 2050 based on 25 years data from the healthcare system in Denmark. We retrieved data from the Danish Hip Arthroplasty Register on 198,835 primary and 29,456 revision surgeries. Historical censuses and population forecasts were retrieved from Statistics Denmark. Logistic and Gompertz regression analysis was used to forecast incidence rates (IR) and total numbers in the next 30 years. Our forecast predicts an increase in IR of 3-9% and an increase in total numbers of primary THA of between 12% and 19% in 2050. For revision THA the IRs have reached a plateau but total numbers are predicted to increase by 19% in 2050. Our forecast shows that both primary and revision THA will increase in total numbers in the next decades, but the IR for primary THA is near its plateau and for revision THA the plateau has already been reached. The forecast may aid in healthcare resource planning for the decades to come.

Supporting weather forecasting performance management at aerodromes through anomaly detection and hierarchical clustering

Weather forecasting is a critical factor for aerodrome and enroute flight operations. Airport decision-makers rely on assessments made by forecasters to ensure operations safety and optimize flight schedule despite potential adverse weather conditions. This manuscript suggests a novel methodology based on Machine Learning to detect forecasting anomalies in historic data, and to rely on them for anticipating potential threats in aerodrome future forecasts. The methodology is fed with historic bulletins from radars and with previous forecasts, which are then processed via an anomaly detection algorithm, and a hierarchical clustering algorithm. While the former algorithm spots anomalous data points, the latter is used to group sets of similar forecasts. The joint usage of the results allows calculating an error propensity metric, which can predict the expected tendency of a certain forecast to be inaccurate. The methodology is meant to enhance decision makers in managing aerodrome weather forecasting, understanding criticalities related to their accuracy levels.

Decision support framework for inventory management combining fuzzy multicriteria methods, genetic algorithm, and artificial neural network

Decision support tools, within the Industry 4.0 perspective, have increasingly impacted different operations and supply chain management (OSCM) areas, such as inventory management. Within the digital transformation era, multicriteria decision-making (MCDM) and machine learning (ML) can be used to improve inventory management decisions. Despite their importance, the literature lacks empirical studies involving advanced solutions that combine both approaches to support practitioners in real-life settings. This is especially relevant for maintenance, repair, and operation (MRO) inventories, which usually present several SKUs with irregular demand patterns and difficult forecasting. Therefore, this study proposes a decision support framework for inventory management, combining MCDM and ML approaches, and applies it to a railway logistics operator to assist its MRO inventory management decision-making process. The first stage of the framework consists of applying a hybrid MCDM method that combines fuzzy logic with the analytic hierarchy process (AHP) and VIseKriterijumska Optimizacija I Kompromisno Resenje (VIKOR) methods to rank and select SKUs according to importance and criticality. Once the most critical SKUs are revealed, a second stage of the framework is introduced to forecast the demand for these SKUs through an ML model, which combines a genetic algorithm and an artificial neural network (GA-ANN). Research findings point to a considerable improvement in the accuracy of the demand forecast for SKUs compared to the previous forecast by the company, and the forecasting methods support vector machine (SVM) and exponential smoothing. The results of the combined approaches are consolidated into a management dashboard, which improves the agility and quality of the analyzed company's decision-making process in inventory management. Therefore, practitioners can also take stock of the proposed framework as a semi-automatic management artefact to enhance decision-making from a digital transformation view in a vital area of OSCM.

Confirmation Bias in Analysts’ Response to Consensus Forecasts

This paper provides evidence of confirmation bias by sell-side analysts in their earnings forecasts. We show that analysts tend to put higher weight on public information when the current forecast consensus is more consistent with their previous forecasts. We further find that analysts with better past forecasting performance, longer firm-specific experience, or forecasting earlier, tend to be more subject to confirmation bias, consistent with some of the existing cognitive and social psychology theories. The results remain significant after controlling for analyst incentives. Finally, we distinguish evidence of confirmation bias from that of other important behavioral biases such as conservative bias.

Pseudo-prospective testing of 5-year earthquake forecasts for California using inlabru

Abstract. Probabilistic earthquake forecasts estimate the likelihood of future earthquakes within a specified time-space-magnitude window and are important because they inform planning of hazard mitigation activities on different time scales. The spatial component of such forecasts, expressed as seismicity models, generally relies upon some combination of past event locations and underlying factors which might affect spatial intensity, such as strain rate, fault location and slip rate or past seismicity. For the first time, we extend previously reported spatial seismicity models, generated using the open source inlabru package, to time-independent earthquake forecasts using California as a case study. The inlabru approach allows the rapid evaluation of point process models which integrate different spatial datasets. We explore how well various candidate forecasts perform compared to observed activity over three contiguous 5-year time periods using the same training window for the input seismicity data. In each case we compare models constructed from both full and declustered earthquake catalogues. In doing this, we compare the use of synthetic catalogue forecasts to the more widely used grid-based approach of previous forecast testing experiments. The simulated catalogue approach uses the full model posteriors to create Bayesian earthquake forecasts, not just the mean. We show that simulated catalogue based forecasts perform better than the grid-based equivalents due to (a) their ability to capture more uncertainty in the model components and (b) the associated relaxation of the Poisson assumption in testing. We demonstrate that the inlabru models perform well overall over various time periods: The full catalogue models perform favourably in the first testing period (2006–2011) while the declustered catalogue models perform better in the 2011–2016 testing period, with both sets of models performing less well in the most recent (2016–2021) testing period. Together, these findings demonstrate a significant improvement in earthquake forecasting is possible although this has yet to be tested and proven in true prospective mode.

Constraining primordial black hole dark matter with CHIME fast radio bursts

Strong lensing of fast radio bursts (FRBs) has been proposed as a relatively clean probe of primordial black hole (PBH) dark matter. Recently, the Canadian Hydrogen Intensity Mapping Experiment (CHIME) published a first catalog of 536 FRBs, 62 of which are from repeating sources. In light of this new data, we reexamine the prospects to constrain the abundance of PBHs via FRBs. Extending previous forecasts, we calculate a PBH dark matter bound using the intrinsic burst width and a calibrated flux-ratio threshold per FRB. In addition, we take into account the uncertainty in the relation between the FRB dispersion measure and source redshift. We outline an algorithm to detect lensed FRBs and a method to simulate its performance on real data and set a flux-ratio threshold for each event, which we use to infer realistic forecasts. We then attempt to extract a preliminary bound using the publicly available CHIME data. Unfortunately, both instrumental noise and the provided $\ensuremath{\sim}1\text{ }\text{ }\mathrm{ms}$ time-resolution of the public data hinder this effort. We identify one candidate event where a double burst could be explained via strong lensing by a $\mathcal{O}(10\text{ }\text{ }{M}_{\ensuremath{\bigodot}})$-mass PBH, which will require follow-up study at higher time resolution to either confirm or discard. We show that with a few times the size of the first catalog---sampled at the full instrumental time-resolution so that candidates can be efficiently scrutinized---CHIME will be able to find strong evidence for or robustly rule out PBHs with mass above $\ensuremath{\sim}10\text{ }\text{ }{M}_{\ensuremath{\bigodot}}$ as the dark matter. Finally, we demonstrate that stacking repeating FRBs can improve the constraints, especially for lower masses.