Forecast Variance Research Articles

The Role of Antecedent Winter Soil Moisture Carryover on Spring Runoff Predictability in Snow-Influenced Western U.S. Catchments

Abstract Spring runoff is critical to agricultural, industrial, and municipal water supply as well as environmental uses in the western United States. Although spring runoff in this region has long been predicted based on late winter and early spring montane snow water storage, the role of soil moisture carryover from the previous winter is less understood and utilized in forecasts. We quantify the relationship between antecedent winter soil moisture and spring runoff for 85 unmanaged catchments in the western United States. Using a regression-based approach, we estimate the proportional reduction in error in seasonal runoff forecasts associated with soil moisture. We classify the catchments into two regimes: interior (cold and relatively dry) and maritime (warmer and wetter), based on seasonal variations in soil moisture. Soil moisture generally is relatively unchanged through the winter period in interior catchments, whereas it increases (mostly gradually, but occasionally rapidly) in maritime catchments. We find that winter temperature dominates soil moisture variability in both types of catchments. We find two patterns in the predictability of spring runoff. First, including antecedent winter soil moisture as a predictor enhances forecast accuracy and variance explanation in spring runoff for both interior and maritime catchments, particularly and with greater significance in interior catchments. Second, spring runoff prediction skill from antecedent winter soil moisture increases with elevation. Overall, we find strong evidence that the use of antecedent soil moisture can improve spring runoff forecast skill for catchments in the western United States. Significance Statement Spring runoff forecasts are crucial for water supply in the western United States, but typically exclude information about soil moisture despite its key role in the water balance. Our study explores how and why incorporating information about winter’s soil moisture levels affects spring runoff forecasts across 85 catchments in the western United States. We find that doing so can produce modest improvements in the prediction skill of spring runoff for both interior and maritime catchments but is particularly noticeable in interior catchments and at higher elevations where temperatures are lower. Overall, our research strongly supports the idea that using information about soil moisture from the previous winter can improve accuracy in spring runoff forecasts for the western United States.

Forecasting U.S. stock returns conditional on geopolitical risk and business cycles

Using standard predictors in the forecasting literature, we forecast the U.S. stock market returns conditional on geopolitical risk and business cycles over the 1927–2021 period. We find that out-of-sample forecasting performance is significantly better in times of high geopolitical risk versus low geopolitical risk. Consistent with previous research, we find further evidence of improved return predictability in recessions. However, we find little difference in forecasting performance in recessions versus expansions once the level of geopolitical risk is controlled for. We find similar results when using stock market cycles and periods of positive/negative industrial production growth in place of recessions/expansions. Our study contributes to the forecasting literature by documenting that geopolitical risk by itself and in combination with business cycle indicators impacts the forecasting ability of standard forecasting variables in the literature. We also contribute to the literature on the adaptive markets hypothesis with evidence of time-varying return predictability. We find inconclusive evidence as to whether our results are based on time-varying predictability or time-varying risk.

Flow dependence of forecast uncertainty in a large convection‐permitting ensemble

AbstractThe flow dependence of forecast distributions is studied using a 120‐member ICON‐D2 ensemble for two representative case studies of weak and strong synoptic forcing of convection. The bootstrapping technique is used to investigate the convergence of sampling error for a range of surface and midtroposphere variables. Convergence is generally observed for the mean and the standard deviation, but not for the 95th percentile, especially in strong forcing. Additionally, maps of uncertainty are introduced, which allow for a more detailed analysis of the spatial pattern of uncertainty and facilitate the interpretation of the sources and evolution of forecast uncertainty in different synoptic forcing conditions. It is found that convection significantly increases uncertainty in model variables and shapes the spatial uncertainty pattern, with weak forcing leading to patchy uncertainty and strong forcing resulting in a more coherent pattern due to large convective systems. Overall, the main result of this work is the strong connection between uncertainty of forecast variables and convection, with synoptic forcing being crucial in determining the spatial distribution and uncertainty evolution within 24 h.

Southern Hemisphere baroclinic activity in seasonal forecasts

Accurate prediction of mid-latitude baroclinic activity is extremely relevant for understanding global climate dynamics and improving long-term weather forecasts. However, current seasonal forecast models struggle to accurately represent the variability of baroclinic activity in the Southern Hemisphere, which may affect their reliability and usefulness. Baroclinic instability in the mid-latitudes is a significant component of the climate system, as it is associated with the meridional transport of a large amount of energy and momentum. Therefore, the ability of the models to correctly predict the properties of the atmospheric circulation in this latitudinal region is a very important requirement. The aim of this study is to estimate the energy of atmospheric phenomena typical of the mid-latitudes, such as baroclinic perturbations, and to understand how seasonal forecasts can be practically used to assess the energy transfer in the atmosphere. We compare the Southern Hemisphere mid-latitude winter variability of the seasonal forecasts of the ECMWF, DWD and Météo France forecasting systems with the ERA5 reanalysis. The analysis is performed by computing the Hayashi spectra of the 500-hPa geopotential height field. Both the reanalysis and the seasonal forecast show a series of peaks in the spectral region of eastward-travelling waves, which corresponds to the high frequency and high wavenumber domain. We quantify the amount of energy released from the atmosphere by calculating the Baroclinic Amplitude Index. The results suggest that seasonal forecasts may not accurately capture the variability of geopotential height power spectra in the Southern Hemisphere, which poses a challenge in correctly distributing the energy over spatial and temporal dimensions. This study will show that this problem is particularly pronounced for wavenumber 4 over a period of 8 days. This misrepresentation likely contributes to the uncertainties in precipitation forecasting, with discrepancies exacerbated by a suboptimal description of baroclinic instability and dynamical components in the models. Our findings highlight the need for an improved representation of baroclinic processes in seasonal forecast models, which could lead to substantial advancements in long-term weather prediction capabilities and in a more complete understanding of climate dynamics.

Scheme using a finite‐volume method for local horizontal derivatives in the spectral model: The non‐hydrostatic case

AbstractIn a typical non‐hydrostatic spectral dynamic numerical weather prediction (NWP) kernel, all forecast variables are transformed between grid point and spectral spaces to compute their gradients and solve the implicit problem. This kernel requires numerous spectral transformations, which depend heavily on extensive global communication and significantly hinder parallel computing efficiency. This paper introduces an innovative non‐hydrostatic spectral kernel that incorporates a finite‐volume method within the spectral framework. We have developed a horizontal divergence (D)‐based structure equation, allowing direct computation of most prognostic variables and their horizontal gradients at grid point space. By doing so, the need for spectral transformations is substantially decreased. Our experiments demonstrate that this new approach reduces the cost of spectral transformations by up to 40%, enhancing the overall model efficiency by 15%–22%. Additionally, a series of tests confirmed the accuracy and stability of this new solver.

Minimization of Forecast Variance Using an Example of ETS Models

Minimization of Forecast Variance Using an Example of ETS Models

FORECASTING ECONOMIC TIME SERIES WITH EXOGENOUS VARIA-BLES: OPPORTUNITIES IN COMPUTATIONAL PARAMETRIZATION TECHNIQUES FOR SUPPORTING PUBLIC DEVELOPMENT POLICIES

This study presents a comprehensive literature review exploring the integration of exogenous variables in economic time series forecasting, focusing on supporting public development policies. It systematically examines various computational parametrization techniques incorporating these external variables, emphasizing their potential to enhance prediction accuracy and address practical challenges in policy-driven contexts. By meticulously analyzing the existing body of research, this review identifies significant gaps and outlines future opportunities for advancing computational techniques and model development. Addressing these gaps could lead to more robust and precise economic forecasting models, offering valuable insights for both academic researchers and policymakers in the formulation and implementation of effective public development strategies. This paper consolidates current knowledge and paves the way for innovative computational approaches that can revolutionize economic forecasting by integrating multidimensional data from diverse external sources.

Higher Education Landscape in India: Government Expenditure and Its Implications on Growth and Access

India’s higher education system grapples with myriad challenges perpetuated by scarcity of resources. This study seeks to analyse the quantitative growth of India’s higher education since independence, particularly focusing on the post-reform period. The research employs regression models, including the semi-log, Gompertz and multiple linear models, to determine growth rates and forecast variables up to 2035. The study reveals a significant expansion of higher education during the examined period, though it falls short of meeting the increasing demands. To achieve the goal of a 50 per cent Gross Enrollment Ratio (GER) by 2035, the Government of India must take steps to boost its expenditure on higher education. The analysis powerfully underscores that despite the extensive proliferation of higher education in India, its efficacy might be limited without concurrent implementation of robust policies directed towards amplifying government expenditure within the sector.

Global Precipitation Nowcasting of Integrated Multi-satellitE Retrievals for GPM: A U-Net Convolutional LSTM Architecture

Abstract This paper presents a deep supervised learning architecture for 30-min global precipitation nowcasts with a 4-h lead time. The architecture follows a U-Net structure with convolutional long short-term memory (ConvLSTM) cells empowered by ConvLSTM-based skip connections to reduce information loss due to the pooling operation. The training uses data from the Integrated Multi-satellitE Retrievals for GPM (IMERG) and a few key drivers of precipitation from the Global Forecast System (GFS). The impacts of different training loss functions, including the mean-squared error (regression) and the focal loss (classification), on the quality of precipitation nowcasts are studied. The results indicate that the regression network performs well in capturing light precipitation (<1.6 mm h−1), while the classification network can outperform the regression counterpart for nowcasting of high-intensity precipitation (>8 mm h−1), in terms of the critical success index (CSI). It is uncovered that including the forecast variables can improve precipitation nowcasting, especially at longer lead times in both networks. Taking IMERG as a relative reference, a multiscale analysis, in terms of fractions skill score (FSS), shows that the nowcasting machine remains skillful for precipitation rate above 1 mm h−1 at the resolution of 10 km compared to 50 km for GFS. For precipitation rates greater than 4 mm h−1, only the classification network remains FSS skillful on scales greater than 50 km within a 2-h lead time. Significance Statement This study presents a deep neural network architecture for global precipitation nowcasting with a 4-h lead time, using sequences of past satellite precipitation data and simulations from a numerical weather prediction model. The results show that the nowcasting machine can improve short-term predictions of high-intensity global precipitation. The research outcomes will enable us to expand our understanding of how modern artificial intelligence can improve the predictability of extreme weather and benefit flood early warning systems for saving lives and properties.

Enhancing Probabilistic Solar PV Forecasting: Integrating the NB-DST Method with Deterministic Models

Accurate quantification of uncertainty in solar photovoltaic (PV) generation forecasts is imperative for the efficient and reliable operation of the power grid. In this paper, a data-driven non-parametric probabilistic method based on the Naïve Bayes (NB) classification algorithm and Dempster–Shafer theory (DST) of evidence is proposed for day-ahead probabilistic PV power forecasting. This NB-DST method extends traditional deterministic solar PV forecasting methods by quantifying the uncertainty of their forecasts by estimating the cumulative distribution functions (CDFs) of their forecast errors and forecast variables. The statistical performance of this method is compared with the analog ensemble method and the persistence ensemble method under three different weather conditions using real-world data. The study results reveal that the proposed NB-DST method coupled with an artificial neural network model outperforms the other methods in that its estimated CDFs have lower spread, higher reliability, and sharper probabilistic forecasts with better accuracy.

Accuracy of historical precipitation from CMIP6 global climate models under diversified climatic features over India

The importance of global climate models (GCMs) is increasingly recognized due to their excellent ability to accurately predict climatic factors. These capabilities prove invaluable to water resources engineers as they facilitate effective planning and strategic decision-making. Finally, evaluating the performance of GCMs is very important because it allows us to simulate and predict different climate scenarios, empowering us to make informed choices. Therefore, the purpose of this study is to determine the degree of discordance between historical simulated data produced by the CMIP6 models and historical observational data over different climate zones of India. The ability of 24 different GCMs to reproduce the geographical and seasonal distribution of Indian precipitation has been tested by analyzing the daily historical precipitation forecasts from these models. These models have been used to estimate the degree of uncertainty associated with the spatiotemporal variability of precipitation forecasts. More than 20% percent bias (PBIAS) is observed to occur predominantly in four climate classifications: polar tundra, temperate, cold, and tropical monsoon. In some regions of India, the CMIP6 models produce overestimated or underestimated results. The locations identified indicate that there have been changes of more than 20% PBIAS near Sivalik Range, Naga Hills, and Western Ghats. The precipitations of those regions that have been underestimated also imply that those locations have different climatic conditions. This study also highlights that CMIP6 GCMs are yet to produce better results near several Indian mountainous regions depending upon climates. The outcomes of this study will be very useful for reconstructing modeled data for that specific regions.

Stock market pattern recognition using symbol entropy analysis

Market uncertainty and asset pricing are closely related. The emergence of global and local shocks provokes price patterns in stock markets that threaten overall market stability. Using a combination of symbol entropy analysis, regression models, and forecast variance decomposition estimates, we examine the relationships between the frequency of states associated with rising and falling price patterns of the S&P 500 index, its implied volatility, and the fluctuations in the global stock market uncertainty. We find a positive (negative) relationship between the increase (decrease) in global uncertainty and the frequency of extreme bear (bull) episodes. Similarly, we observe the same relationship between increases (decreases) in the implied volatility and the frequency of extreme bear (bull) episodes. Furthermore, joint increases (decreases) in implied volatility and global uncertainty are positively (negatively) associated with the frequency of bear (bull) episodes in the S&P 500 index. Finally, interconnectedness studies confirm the relevance of the mutual influence of the phenomenon and its dynamic nature, which increases the probability of observing shock waves of financial instability in global and local stock markets. Our work highlights the need to consider price patterns and their links to uncertainty as factors of stock market instability. For policymakers, governments, and practitioners, our findings are a call to include events with high global/local impact in the radar of threats to financial stability.

Focal-TSMP: deep learning for vegetation health prediction and agricultural drought assessment from a regional climate simulation

Abstract. Satellite-derived agricultural drought indices can provide a complementary perspective of terrestrial vegetation trends. In addition, their integration for drought assessments under future climates is beneficial for providing more comprehensive assessments. However, satellite-derived drought indices are only available for the Earth observation era. In this study, we aim to improve the agricultural drought assessments under future climate change by applying deep learning (DL) to predict satellite-derived vegetation indices from a regional climate simulation. The simulation is produced by the Terrestrial Systems Modeling Platform (TSMP) and performed in a free evolution mode over Europe. TSMP simulations incorporate variables from underground to the top of the atmosphere (ground-to-atmosphere; G2A) and are widely used for research studies related to water cycle and climate change. We leverage these simulations for long-term forecasting and DL to map the forecast variables into normalized difference vegetation index (NDVI) and brightness temperature (BT) images that are not part of the simulation model. These predicted images are then used to derive different vegetation and agricultural drought indices, namely NDVI anomaly, BT anomaly, vegetation condition index (VCI), thermal condition index (TCI), and vegetation health index (VHI). The developed DL model could be integrated with data assimilation and used for downstream tasks, i.e., for estimating the NDVI and BT for periods where no satellite data are available and for modeling the impact of extreme events on vegetation responses with different climate change scenarios. Moreover, our study could be used as a complementary evaluation framework for TSMP-based climate change simulations. To ensure reliability and to assess the model’s applicability to different seasons and regions, we provide an analysis of model biases and uncertainties across different regions over the pan-European domain. We further provide an analysis about the contribution of the input variables from the TSMP model components to ensure a better understanding of the model prediction. A comprehensive evaluation of the long-term TSMP simulation using reference remote sensing data showed sufficiently good agreements between the model predictions and observations. While model performance varies on the test set between different climate regions, it achieves a mean absolute error (MAE) of 0.027 and 1.90 K with coefficient of determination (R2) scores of 0.88 and 0.92 for the NDVI and BT, respectively, at 0.11° resolution for sub-seasonal predictions. In summary, we demonstrate the feasibility of using DL on a TSMP simulation to synthesize NDVI and BT satellite images, which can be used for agricultural drought forecasting. Our implementation is publicly available at the project page (https://hakamshams.github.io/Focal-TSMP, last access: 4 April 2024).

Investigation of Load, Solar and Wind Generation as Target Variables in LSTM Time Series Forecasting, Using Exogenous Weather Variables

Accurately forecasting energy metrics is essential for efficiently managing renewable energy generation. Given the high variability in load and renewable energy power output, this represents a crucial area of research in order to pave the way for increased adoption of low-carbon energy solutions. Whilst the impact of different neural network architectures and algorithmic approaches has been researched extensively, the impact of utilising additional weather variables in forecasts have received far less attention. This article demonstrates that weather variables can have a significant influence on energy forecasting and presents methodologies for using these variables within a long short-term memory (LSTM) architecture to achieve improvements in forecasting accuracy. Moreover, we introduce the use of the seasonal components of the target time series, as exogenous variables, that are also observed to increase accuracy. Load, solar and wind generation time series were forecast one hour ahead using an LSTM architecture. Time series data were collected in five Spanish cities and aggregated for analysis, alongside five exogenous weather variables, also recorded in Spain. A variety of LSTM architectures and hyperparameters were investigated. By tuning exogenous weather variables, a 33% decrease in mean squared error was observed for solar generation forecasting. A 22% decrease in mean absolute squared error (MASE), compared to 24-h ahead forecasts made by the Transmission Service Operator (TSO) in Spain, was also observed for solar generation. Compared to using the target variable in isolation, utilising exogenous weather variables decreased MASE by approximately 10%, 15% and 12% for load, solar and wind generation, respectively. By using the seasonal component of the target variables as an exogenous variable itself, we demonstrated decreases in MASE of 19%, 12% and 8% for load, solar and wind generation, respectively. These results emphasise the significant benefits of incorporating weather and seasonal components into energy-related time series forecasts.

Visualizing the Shelf Life of Population Forecasts: A Simple Approach to Communicating Forecast Uncertainty

It is widely appreciated that population forecasts are inherently uncertain. Researchers have responded by quantifying uncertainty using probabilistic forecasting methods. Yet despite several decades of development, probabilistic forecasts have gained little traction outside the academic sector. Therefore, this article suggests an alternative and simpler approach to estimating and communicating uncertainty which might be helpful for population forecast practitioners and users. Drawing on the naïve forecasts idea of Alho, it suggests creating “synthetic historical forecast errors” by running a regular deterministic projection model many times over recent decades. Then, borrowing from perishable food terminology, the “shelf life” of forecast variables, the number of years into the future the forecast is likely to remain “safe for consumption” (within a specified error tolerance), is estimated from the “historical” errors. The shelf lives are then applied to a current set of forecasts and presented in a simple manner in graphs and tables of forecasts using color-coding. The approach is illustrated through a case study of 2021-based population forecasts for Australia. It´s concluded that the approach offers a relatively straightforward way of estimating and communicating population forecast uncertainty.

Memory of HF radar surface-current assimilation on different forecasting variables of ocean circulation models

In this study, using nudging method, the current data from a pair of HF radar along Jiangsu coast was assimilated to unstructured grid models based on FVCOM. Several experiments were conducted to examine the effect of radar data on surface velocity, elevation and sea surface salinity forecast. The results show that the gain of assimilation on surface velocity and elevation forecasts disappeared fast within a few hours, while salinity forecasts can be influenced for more than 10 days. This fast diminishing of the gain of data assimilation on current and elevation can be attributed to the fast wave propagation from elsewhere to the radar coverage area.

Analyst forecasting and market liquidity dynamics: A comparative analysis of U.S. and Chinese markets pre and post COVID-19

This study investigates the influence of analyst forecasting on market liquidity and information efficiency within developed U.S. and emerging Chinese markets, both before and during the global COVID-19 pandemic. Market liquidity, a crucial aspect of financial market functioning, is assessed using key indicators, including Amihud’s liquidity, turnover and bid/ask spreads. The findings reveal notable disparities in analysts’ forecasts of market dynamics before and after the onset of the COVID-19 crisis. During the pre-COVID-19 period, our analysis confirms a statistically significant relationship between analyst forecasting and market liquidity. However, as the COVID-19 pandemic emerged and evolved, analysts’ forecasting impacts on market liquidity showed a diminishing trend. Furthermore, the effects of analyst forecasting vary across quantiles in the Chinese market. In the United States, buy-sell recommendations and various analyst forecasting variables were significant before the pandemic. These findings underscore the nuanced interplay between analysts’ forecasts and the market dynamics within the studied regions. Most notably, the U.S. and Chinese markets experienced decreased information efficiency during the COVID-19 crisis. This shift resulted in a notable decoupling of analysts’ forecasts from information efficiency during this period. These outcomes have significant implications for scholars and investors alike, enhancing their comprehension of market conditions and investor behaviour when confronted with heightened market stressors, such as the COVID-19 pandemic, which has disrupted international financial markets.

Prediction of distributions of unit prices for real estate properties on the basis of the characteristics of PSI-processes

Real estate market price forecasting is always in the focus of interests of scientists-economists, market analysts, market participants (sellers and buyers), marketing services of building complex enterprises, analysts working for banks and insurance companies and investors. Under present day conditions, the price behavior of properties on real estate markets takes especially important meaning subject to the influence of such factors as changes in the structure of household incomes, changes in mortgage rates and their availability, dynamic changes in the macroeconomic and other external socio-economic and political type factors. However, unlike the financial and securities markets, the real estate market is always characterized by a delayed reaction to external perturbations, often up to half a year, which allows us to hope for an appropriate construction of forecasts, at least in time for the delayed reaction. Traditional autoregressive forecasting methods are characterized by rapidly increasing forecast variance, because they assume a factor of stochastic volatility. This paper proposes a model and method of forecast construction based on stochastic processes of the “Poisson random index” having a short time for reaching a stationary stable variance. The model is based on the “principle of replacements” of current prices with new ones. We analyze in detail an example of the application of the “principle of replacements” for construction of price forecasts on secondary residential real estate in St. Petersburg which is based on data of four-year observations of offer prices.

Application of AI for Short-Term PV Generation Forecast.

The efficient use of the photovoltaic power requires a good estimation of the PV generation. That is why the use of good techniques for forecast is necessary. In this research paper, Long Short-Term Memory, Bidirectional Long Short-Term Memory and the Temporal convolutional network are studied in depth to forecast the photovoltaic power, voltage and efficiency of a 1320 Wp amorphous plant installed in the Technology Support Centre in the University Rey Juan Carlos, Madrid (Spain). The accuracy of these techniques are compared using experimental data along one year, applying 1 timestep or 15 min and 96 step times or 24 h, showing that TCN exhibits outstanding performance, compared with the two other techniques. For instance, it presents better results in all forecast variables and both forecast horizons, achieving an overall Mean Squared Error (MSE) of 0.0024 for 15 min forecasts and 0.0058 for 24 h forecasts. In addition, the sensitivity analyses for the TCN technique is performed and shows that the accuracy is reduced as the forecast horizon increases and that the 6 months of dataset is sufficient to obtain an adequate result with an MSE value of 0.0080 and a coefficient of determination of 0.90 in the worst scenarios (24 h of forecast).

The effect of electricity consumption determinants in household load forecasting models

Usually, household electricity consumption fluctuates, often driven by several electrical consumption determinants such as income, household size, and price. Recently, research studies on the investigation of predictor variables in household electricity consumption have increased especially in the developing and newly industrialized countries. However, the studies just focus on identifying the predictor variables of household electricity consumption that influence load forecasting models. In Tanzania, for instance, scholars found that using the “income” determinant improves the performance of a forecasting model. The scholars suggest without any empirical bases that adding more predictor variables would have improved the accuracy of the model. This study aims to analyze the effect of the number of predictor variables on household load forecasting performance based on Tanzania’s data. Nonlinear regression based on a Weibull function and multivariate adaptive regression splines approaches are used for this purpose. Our findings indicate that income, household size, and number of appliances are common predictor variables of household consumption in developing countries. The measured forecasting root-mean-square error (RMSE) when using income, household size, and the number of appliances is 0.8244, 1.2314, and 0.9868, respectively. Finally, we forecasted load using all three determinants and the RMSE dropped to 0.7031. Having obtained the smaller value of RMSE when all predictors are used reveals that the inclusion of all three predictor variables in load forecasting leads to a significant decrease in RMSE by 14.73%. Therefore, the study recommends using multiple predictor variables in load forecasting models to increase accuracy.