Least-squares Estimation Research Articles

A detailed analysis of stochastic models applied to temporal gravity field recovery with GRACE observations

Summary In this study, we analyzed the impacts of errors in background force models and observed non-gravitational forces on the pseudo-observations (pre-fits) during gravity field recovery based on the GRACE satellite gravity mission. To reduce these effects, we introduced the stochastic parameters into the functional model of the variational equation integration approach to absorb this type of noise contribution. Simultaneously, the prior variances of observed orbits and K-Band range rates used in traditional method are re-estimated with least-square variance component estimation (LS-VCE) after considering these stochastic parameters. To improve the computing efficiency, a modified method of the calculation of sensitivity matrices related to the introduced stochastic parameters is proposed. Compared to the method of variation of constants widely used in the precise orbit determination and gravity field recovery, the modified method decreases the computational time of these matrices by about four times. Furthermore, an efficient LS-VCE algorithm is derived in a more generalized case. The efficient algorithm only costs 1% of the time of the unoptimized method. With the GRACE data, we analyzed the benefits of these refinements in gravity field recovery, and the results show that these improvements can mitigate the impacts of errors in background force models and accelerometer data on recovered gravity field models, especially in the high degree signals. Furthermore, the quality of results has less dependence on parameterization.

Estimation and variable selection for high-dimensional spatial dynamic panel data models

Spatiotemporal modeling of networks is of great practical importance, with modern applications in epidemiology and social network analysis. Despite rapid methodological advances, how to effectively and efficiently estimate the parameters of spatial dynamic panel models remains a challenging problem. To tackle this problem, we construct consistent complex least-squares estimators by the eigendecomposition of a spatial weight matrix method originally proposed for undirected networks. We no longer require all eigenvalues and eigenvectors to be real, which is a remarkable achievement as it implies that the proposed method is now applicable to spatiotemporal data modeling of directed networks. Under mild, interpretable conditions, we show that the proposed parameter estimators are consistent and asymptotically normally distributed. We also present a complex orthogonal greedy algorithm for variable selection and rigorously investigate its convergence properties. Moreover, we incorporate fixed effects into the spatial dynamic panel models and provide a model transformation so that the proposed method can also be applied to the transformed model. Extensive simulation studies and data examples demonstrate the effectiveness of the proposed method.

Estimating Optimal Cost, Insulation Layer Thickness, and Structural Layer Thickness of Different Composite Insulation External Walls Using Computational Methods

All the modern gadgets and space conditioning in buildings consume lots of energy. Energy consumption can be optimized using Composite Insulation External Walls (CIEW) built from mortar plaster and structural and insulation layers. This study aimed to improve the overall performance of CIEW by optimizing the structural and insulation layer thickness. The objective was to minimize the Life Cycle Cost (LCC) and maximize the Life Cycle Savings (LCS) of CIEW. The nonlinear Least Squares Estimation (LSE) optimization technique for optimizing LCC and LCS of CIEW was used in the study. The study considered three insulation materials—Extruded Polystyrene (XPS), Rock Wool (RW), and Glass Wool (GW)—across three heat sources, including Circulating Fluidized Bed (CFB), Grate-Fired Boiler (GFB), and Air-Source Heat Pump (ASHP). The Life Cycle Cost Analysis (LCCA) methodology suggested by Huang using a traditional optimization technique was used as a basis for mathematical formulations and result comparison. The payback period of CIEW with optimal structural and insulation layer thickness was computed. The findings revealed that applying the LSE method enabled greater economic efficiency than the LCCA method, with an up to 9.12% increase in LCS value and an up to 7.41% decrease in LCC value. The research also revealed significant correlations between insulation and structural layer thicknesses and economic parameters.

Unifying Estimation and Inference for Linear Regression with Stationary and Integrated or Near-Integrated Variables

Abstract There is a discrepancy in the limiting distributions of least-squares estimators for stationary and integrated variables. For statistical inference, it must be decided which distribution should be used in advance. This motivates us to develop a unifying inference procedure based on weighted estimation. The asymptotic distributions of the proposed estimators are developed and a random weighting bootstrap method is proposed for constructing confidence regions. The proposed method outperforms existing methods (with time constant or time-varying error variance) in simulations. We further study the predictability of asset returns in a setting where some of our state variables are endogenous.

The impact of ownership structure and board of directors on performance, financing, environmental, social, and governance

Subject. The article addresses the features of agency relations, i.e. concentration, ownership structure and Board of Directors that act as non-financial factors affecting the performance of companies. Objectives. The study aims to substantiate and empirically establish linkages between the specifics of agency relations, as non-financial drivers, with company’s performance and responsibility, measured through ESG. Methods. The use of methods of econometric analysis and least-squares estimation determine the links and strength of the impact of non-financial factors on return, Tobin’s Q, debt-to-asset ratio, and ESG. Results. For Russian companies, the paper underpins and evaluates the impact of concentration on an increase in return on assets and equity, but a decrease in value. Government ownership decreases performance, but raises financing. The structure of the Board of Directors impacts an increase in ROA, ROE, and Tobin’s Q, and a decrease in debt-to-asset ratio. For some Russian, U.S., European and Japanese companies, the paper empirically proved a stimulating effect of the size and presence of women in the Board of Directors on ESG. Decomposition of ESG refines the results. The size of the Board of Directors to a greater extent improves Environmental for American, European companies, and independent directors – Environmental for Russian companies. The presence of women improves Governance for American and European firms, and Social – for Russian and Japanese companies. Conclusions. The findings enhance understanding of agency relations with company performance, value, and fund raising. Conclusions about ESG dependence on the Board of Directors structure and size are applied in the context of the Sustainable Development Goals.

Machine learning-based radiomics to differentiate immune-mediated necrotizing myopathy from limb-girdle muscular dystrophy R2 using MRI.

This study aimed to assess the feasibility of a machine learning-based radiomics tools to discriminate between Limb-girdle muscular dystrophy R2 (LGMDR2) and immune-mediated necrotizing myopathy (IMNM) using lower-limb muscle magnetic resonance imaging (MRI) examination. After institutional review board approval, 30 patients with genetically proven LGMDR2 (12 females; age, 34.0 ± 11.3) and 45 patients with IMNM (28 females; age, 49.2 ± 16.6) who underwent lower-limb MRI examination including T1-weighted and interactive decomposition water and fat with echos asymmetric and least-squares estimation (IDEAL) sequences between July 2014 and August 2022 were included. Radiomics features of muscles were obtained, and four machine learning algorithms were conducted to select the optimal radiomics classifier for differential diagnosis. This selected algorithm was performed to construct the T1-weighted (TM), water-only (WM), or the combined model (CM) for calf-only, thigh-only, or the calf and thigh MR images, respectively. And their diagnostic performance was studied using area under the curve (AUC) and compared to the semi-quantitative model constructed by the modified Mercuri scale of calf and thigh muscles scored by two radiologists specialized in musculoskeletal imaging. The logistic regression (LR) model was the optimal radiomics model. The performance of the WM and CM for thigh-only images (AUC 0.893, 0.913) was better than those for calf-only images (AUC 0.846, 0.880) except the TM. For "calf + thigh" images, the TM, WM, and CM models always performed best (AUC 0.953, 0.907, 0.953) with excellent accuracy (92.0, 84.0, 88.0%). The AUCs of the Mercuri model of the calf, thigh, and "calf + thigh" images were 0.847, 0.900, and 0.953 with accuracy (84.0, 84.0, 88.0%). Machine learning-based radiomics models can differentiate LGMDR2 from IMNM, performing better than visual assessment. The model built by combining calf and thigh images presents excellent diagnostic efficiency.

Determinants of Cash Distribution Options in South African Listed Firms: An Empirical Analysis of Earnings, Company Size, and Economic Value Added

The purpose of this study was to examine the determinants of cash distribution options by critically considering the effects of earnings, dividends, firm size, and economic value added. The distribution of cash dividends to shareholders serves as a basic means by which shareholders receive returns on their investments, so it is essential to examine share repurchases alongside dividends to enhance management’s efforts in maximising shareholder value. This study utilised panel data from 52 companies listed on the Johannesburg Security Exchange (JSE) that engaged in open market share repurchases for at least 2 years between 2000 and 2019. The data were extracted from the IRESS database. The panel data regression model was fitted with the ordinary least squares (OLS), difference generalised moment method (Diff-GMM), system generalised moment method (Sys-GMM), and least-squares dummy variable correction estimator (LSDVC). The findings revealed that there was a positive and significant relationship between the earnings per share and the payoff flexibility, implying that there was an inherent flexibility of repurchases as a payout option in the sampled firms. Additionally, the study revealed a significant negative relationship between the firm size, economic value added, and payoff flexibility. This suggests that larger companies tend to distribute a lower proportion of their earnings as share repurchases and opt for higher cash dividends instead. The implications of these findings provide financial managers with valuable insights into the role of share repurchases as a cash distribution choice. By recognising share repurchases as a viable option, financial managers can enhance their efforts to create and maximise shareholder value, particularly in emerging market settings. This evidence should encourage financial managers to recognise share repurchases more as a distribution choice, diffusing the tension regarding share repurchases replacing the payment of cash dividends and some doubt that they may not possess attributes complimentary to cash dividends. The study recommended relevant academic, industry, and policy implications in the South African context.

Outliers in diffusion-weighted MRI: Exploring detection models and mitigation strategies

Diffusion-weighted MRI (dMRI) is a medical imaging method that can be used to investigate the brain microstructure and structural connections between different brain regions. The method, however, requires relatively complex data processing frameworks and analysis pipelines. Many of these approaches are vulnerable to signal dropout artefacts that can originate from subjects moving their head during the scan.To combat these artefacts and eliminate such outliers, researchers have proposed two approaches: to replace outliers or to downweight outliers during modelling and analysis. With the rising interest in dMRI for clinical research, these types of corrections are increasingly important. Therefore, we set out to investigate the differences between outlier replacement and weighting approaches to help the dMRI community to select the best tool for their data processing pipelines. We evaluated dMRI motion correction registration and single tensor model fit pipelines using Gaussian Process and Spherical Harmonic based replacement approaches and outlier downweighting using highly realistic whole-brain simulations. As a proof of concept, we applied these approaches to dMRI infant data sets that contained varying numbers of dropout artefacts.Based on our results, we concluded that the Gaussian Process based outlier replacement provided similar tensor fit results to Gaussian Process based outlier detection and downweighting. Therefore, if only the least-squares estimate of the single tensor model is of interest, our recommendation is to use outlier replacement. However, outlier downweighting can potentially provide a more accurate estimate of the model precision which could be relevant for applications such as probabilistic tractoraphy.

A Cautionary Note on the Use of Linear Regression for Hypothesis Testing

This note draws researchers’ attention to the use of linear regression for the purpose of conducting a hypothesis testing. Even when multiple explanatory variables are included in a regression equation to preclude the hazard of a simple regression of omitting other factors, multicollinearity unfortunately is inherent in multiple regression simply because the included explanatory variables can share some common parameter domain; that is, they co-vary. Here we shall show that however one transforms the variance-covariance matrix of the least-squares estimation to reduce the estimation errors, the procedure amounts to affine transformations of the explanatory variables so that despite the transformation they remain to co-vary, rendering the coefficient as a partial derivative invalid. The root of this problem originates from the fact that one receives the explanatory variables’ values by observation rather than predetermining their values and then collecting the corresponding dependent variable’ values. This situation becomes especially disconcerting when a transformed explanatory variable has its estimated coefficient enjoying an exceptional degree of confidence but of no mathematical status as a partial derivative, misleading engineering or medical prescriptions and public policies.

Real-time water/fat imaging at 0.55T with spiral out-in-out-in sampling.

To develop an efficient and flexible water/fat separated real-time MRI (RT-MRI) method using spiral out-in-out-in (OIOI) sampling and balanced SSFP (bSSFP) at 0.55T. A bSSFP sequence with golden-angle spiral OIOI readout was developed, capturing three echoes to allow water/fat separation. A low-latency reconstruction that combines all echoes was available for online visualization. An offline reconstruction provided water and fat RT-MRI in two steps: (1) image reconstruction with spatiotemporally constrained reconstruction (STCR) and (2) water/fat separation with hierarchical iterative decomposition of water and fat with echo asymmetry and least-squares estimation (HIDEAL). In healthy volunteers, spiral OIOI was acquired in the wrist during a radial-to-ulnar deviation maneuver, in the heart without breath-hold and cardiac gating, and in the lower abdomen during free-breathing for visualizing small bowel motility. We demonstrate successful water/fat separated RT-MRI for all tested applications. In the wrist, resulting images provided clear depiction of ligament gaps and their interactions during the radial-to-ulnar deviation maneuver. In the heart, water/fat RT-MRI depicted epicardial fat, provided improved delineation of epicardial coronary arteries, and provided high blood-myocardial contrast for ventricular function assessment. In the abdomen, water-only RT-MRI captured small bowel mobility clearly with improved water-fat contrast. We have demonstrated a novel and flexible bSSFP spiral OIOI sequence at 0.55T that can provide water/fat separated RT-MRI with a variety of application-specific temporal resolution and spatial resolution requirements.

Algorithms and statistical analysis for linear structured weighted total least squares problem

Weighted total least squares (WTLS) have been regarded as the standard tool for the errors-in-variables (EIV) model in which all the elements in the observation vector and the coefficient matrix are contaminated with random errors. However, in many geodetic applications, some elements are error-free and some random observations appear repeatedly in different positions in the augmented coefficient matrix. It is called the linear structured EIV (LSEIV) model. Two kinds of methods are proposed for the LSEIV model from functional and stochastic modifications. On the one hand, the functional part of the LSEIV model is modified into the errors-in-observations (EIO) model. On the other hand, the stochastic model is modified by applying the Moore-Penrose inverse of the cofactor matrix. The algorithms are derived through the Lagrange multipliers method and linear approximation. The estimation principles and iterative formula of the parameters are proven to be consistent. The first-order approximate variance-covariance matrix (VCM) of the parameters is also derived. A numerical example is given to compare the performances of our proposed three algorithms with the STLS approach. Afterwards, the least squares (LS), total least squares (TLS) and linear structured weighted total least squares (LSWTLS) solutions are compared and the accuracy evaluation formula is proven to be feasible and effective. Finally, the LSWTLS is applied to the field of deformation analysis, which yields a better result than the traditional LS and TLS estimations.

Monitoring of recently sea level changes on the coast of New Zealand using XTRACK coastal altimetry and tide gauge data

The rise in sea level along the coasts of New Zealand has accelerated in recent decades due to the impact of climate change. Determining the effects of these changes on the coastal regions is critical for their sustainability. In order to gain insight into these impacts, the present study aimed to analyze sea level changes using satellite altimetry and tide gauge data from 11 tide gauge stations along the New Zealand coast between 1993 and 2022, as well as XTRACK data processed with the coastal altimetry approach to minimize the effects of classical altimetry near coasts. The least-squares parameter estimation method was used to analyze the data and determine annual and semi-annual seasonal effects. The results showed that a significant increase have been observed in sea level of 3.8+0.5 mm/yr along the coast of New Zealand from 1993 to 2022 using coastal altimetry data. The results of this study demonstrate the importance of closely monitoring the impacts of sea level rise on the coastal regions of New Zealand to ensure their long-term sustainability. The results also highlight the utility of using multiple data sources and methods to provide a comprehensive understanding of these changes.

Long-term outcomes of laparoscopic sleeve gastrectomy in those with class I obesity: safety, efficacy, and quality of life.

Class I obesity carries significant morbidity and mortality risk similar to higher grades of obesity, and persons with class I obesity have a high risk of progression to class II and III obesity. While bariatric surgery has made strides in safety and efficacy, it remains inaccessible for persons with class I obesity (body mass index [BMI] of 30-35 kg/m2). To assess safety, weight loss durability, co-morbidity resolution, and quality of life after laparoscopic sleeve gastrectomy (LSG) in persons with class I obesity. Multidisciplinary medical center that specializes in obesity management. A longitudinal prospective single-surgeon registry was queried for data pertaining to persons with class I obesity who underwent primary LSG. Primary endpoint was weight loss. Secondary endpoints included change in obesity-related co-morbidities, adverse events, and post hoc analysis of symptoms of gastroesophageal reflux disease (GERD) and Bariatric Analysis and Reporting Outcome System results. Follow-up was divided into short- (1-3 yr), intermediate- (4-7 yr), and long-term (8-12 yr). We evaluated percent excess weight loss (%EWL) using linear mixed models adjusting for age, sex, years since operation, and baseline BMI. Least-squares means estimates and 95% confidence intervals (CI) were generated. Of 13,863 bariatric procedures, a total of 1851 patients were included. Mean baseline BMI, age, and male:female ratio were 32.6 ± 2.1 kg/m2, 33.7 ± 9.2 years, and 1:5, respectively. Adjusted mean %EWL (95% CI) at short-, intermediate-, and long-term follow-up were 111% (95% CI, 91%-131%), 110% (95% CI, 89%-131%), and 141% (95% CI, 57%-225%), respectively. Of 195 patients with type 2 diabetes, 59% experienced complete remission, and of 168 patients with hypertension, 43% experienced complete remission. Being on oral antidiabetes medication was a significant predictor of sustained remission compared with being on insulin or combination therapy (P < .001). Sixty-nine patients had symptoms of GERD before surgery, which improved in 55 (79.7%). Thirty-three patients developed de novo symptoms of GERD. The average Bariatric Analysis and Reporting Outcome System score was 4.5 ± 1.7, with 83% of participants rating their quality of life after surgery as good, very good, or excellent. Those with class I obesity who undergo LSG experience normalization of weight, sustained remission of co-morbidity, and good quality of life without significant risk of morbidity or mortality.

Tidal harmonics retrieval using GNSS-R dual-frequency complex observations

Tidal analysis and methods for estimation and prediction of ocean tidal constitutes are essential in a large area of scientific disciplines, for example, navigation, onshore and offshore engineering, and production of green energy. Ground-based Global Navigation Satellite System-Reflectometry (GNSS-R) has been proposed as an alternative method for measuring sea surface height. We use 6 years of GNSS-R observations at In-phase and Quadrature levels from July 2015 to May 2021 obtained from a dedicated receiver and sea-looking left hand circular polarization antenna for estimating sea level (SL). In the first step, the multivariate least-square harmonic estimation (LS-HE) method is applied for SL estimation. Then, final SL time series are generated by combining estimated SL from all satellites at L1 and L2 frequencies in the averaging step. The 6-year root-mean-square error between GNSS-R L12 sea surface heights and a collocated tide gauge (TG) is 5.8 cm with a correlation of 0.948 for a high temporal resolution of 5 min with 15 min averaging window. Afterward, using the univariate LS-HE, we detect tidal harmonics with periods between 30 min to 1 year. The detection results highlight a good match between GNSS-R and TG. Higher harmonics, i.e., the periods shorter than 3 h, show stronger signatures in GNSS-R data. Finally, we estimate the amplitude and phase of standard tidal harmonics from the two datasets. The results show an overall good agreement between the datasets with a few exceptions.

An integrated approach to estimate storage reliability with masked data from series system

Storage reliability is of importance for the products that largely stay in storage in their total life-cycle such as warning systems for harmful radiation detection, etc. Usually, the field-testing data can be available, but the failure causes for a series system cannot be always known because of the masked information. In this paper, the storage reliability model with possibly initial failures is studied on the statistical analysis method when the masked data are considered. To optimize the use of the masked survival data from storage systems, a technique based on the least squares (LS) method with an EM-like algorithm, is proposed for the series system. The parametric estimation procedure based on the LS method is developed by applying the algorithm to update the testing data, and then the LS estimation for the initial reliability and failure rate of the components constituting the series system are investigated. The results should be useful for accurately evaluating the production reliability, identifying the production quality, and planning a storage environment.

Change-Point Detection in the Volatility of Conditional Heteroscedastic Autoregressive Nonlinear Models

This paper studies single change-point detection in the volatility of a class of parametric conditional heteroscedastic autoregressive nonlinear (CHARN) models. The conditional least-squares (CLS) estimators of the parameters are defined and are proved to be consistent. A Kolmogorov–Smirnov type-test for change-point detection is constructed and its null distribution is provided. An estimator of the change-point location is defined. Its consistency and its limiting distribution are studied in detail. A simulation experiment is carried out to assess the performance of the results, which are compared to recent results and applied to two sets of real data.

End-to-end DVB-S2X system design with DL-based channel estimation over satellite fading channels at Ka-band

Satellite channels suffer from heavy fading due to the atmospheric impairments at high frequencies. Therefore, channel estimation is essential for coherent detection and demodulation in satellite-coded systems. The conventional minimum mean square error (MMSE) estimator (the theoretical upper bound) requires a priori knowledge about the channel statistics which is not feasible to obtain in a real transmission. Also, it suffers from high complexity. However, deep learning (DL) estimators do not require any information about the channel statistics. Therefore, in this paper, two DL-based channel estimators are proposed for digital and video broadcasting second generation extension (DVB-S2X) system with less complexity than the MMSE estimator. In particular, the bidirectional long-short term memory (BLSTM) and the gated recurrent unit (GRU) are adopted in our proposed estimators which are termed as DLBLSTM and DLGRU, respectively. The proposed estimators evaluated over two satellite fading channels; one with heavy fading and the other with low fading. Moreover, these estimators are compared with the conventional estimators, the least square (LS) and the MMSE, in terms of the normalized mean square error (NMSE) and the bit error rate (BER). Simulation results show that the proposed DL-based estimators have better performance than the LS estimator and the DLBLSTM has better performance than the DLGRU estimator in terms of NMSE and BER with both satellite channels.

Efficient calibration for imperfect epidemic models with applications to the analysis of COVID-19

Abstract The estimation of unknown parameters in simulations, also known as calibration, is crucial for practical management of epidemics and prediction of pandemic risk. A simple yet widely used approach is to estimate the parameters by minimising the sum of the squared distances between actual observations and simulation outputs. It is shown in this paper that this method is inefficient, particularly when the epidemic models are developed based on certain simplifications of reality, also known as imperfect models which are commonly used in practice. To address this issue, a new estimator is introduced that is asymptotically consistent, has a smaller estimation variance than the least-squares estimator, and achieves the semiparametric efficiency. Numerical studies are performed to examine the finite sample performance. The proposed method is applied to the analysis of the COVID-19 pandemic for 20 countries based on the susceptible-exposed-infectious-recovered model with both deterministic and stochastic simulations. The estimation of the parameters, including the basic reproduction number and the average incubation period, reveal the risk of disease outbreaks in each country and provide insights to the design of public health interventions.

Online state‐of‐charge estimation by modified Coulomb counting method based on the estimated parameters of lithium‐ion battery

SummaryThe state‐of‐charge (SOC) of the battery is the most critical parameter for an electric vehicle (EV) dashboard. Generally, the Coulomb counting (CC) method is used for SOC estimation, but the conventional CC method has some errors due to initial SOC determination. This work uses battery modeling and parameter estimation of a lithium‐iron‐phosphate (LFP) battery to determine the initial SOC. The equivalent circuit models (ECMs) have been used for battery parameters estimation by the least‐squares estimation (LSE) algorithm. This algorithm uses only experimental terminal voltage and current data from Arbin battery testing system. The ECM parameters vary with battery SOC, and the differential capacity plot has explained the variation. SOC can be estimated using estimated parameters and the relationship between SOC and open‐circuit voltage (OCV). Additionally, a modified CC method is proposed to solve the problem of estimating initial SOC by determining an OCV‐SOC relation using experimental data. The modeling of the battery is validated by comparing ECMs with physical battery terminal voltage profiles.

Mapping site index in coniferous forests using bi-temporal airborne laser scanning data and field data from the Swedish national forest inventory

Recent advancements in remote sensing of forests have demonstrated the capabilities of three-dimensional data acquired by airborne laser scanning (ALS) and, consequently, have become an integral part of enhanced forest inventories in Northern Europe. In Sweden, the first national laser scanning revolutionised forest management planning through low-cost production of large-scale and spatially explicit maps of forest attributes such as basal area, volume, and biomass, compared to the earlier practice based on field survey data. A second scanning at the national level was launched in 2019, and it provides conditions for the estimation of height growth and site index. Accurate and up-to-date information about site productivity is relevant for planning silvicultural treatments and for the prognosis of forest status and development over time. In this study, we explored the potential of bi-temporal ALS data and other auxiliary information to predict and map site productivity by site index according to site properties (SIS) of Norway spruce (Picea abies (L.) Karst) and Scots pine (Pinus sylvestris L.) in even-aged stands in Sweden. We linked ground survey data of SIS from more than 11,500 plots of the Swedish National Forest Inventory (NFI) to bi-temporal ALS data to predict and map site index using an area-based method and two regression modelling strategies: (1) a multiple linear regression (MLR) model with an ordinary least-squares parameter estimation method, and (2) a non-parametric random forests (RF) model optimised for hyper-parameter tuning. For model development, permanent plots were used, whereas the validation was done on the temporary plots of the Swedish NFI and an independent stand-level dataset. Species-specific models were developed, and the root mean square error (RMSE) metric was used to quantify the residual variability around model predictions. For both species, the MLR model gave precise and accurate estimates of SIS. The RMSE for SIS predictions was in the range of 1.96 – 2.11 m, and the relative RMSE was less than 10 % (7.68 – 9.49 %) of the reference mean value. Final predictors of site index include metrics of 90th percentile height and annual increment in the 95th percentile height, altitude, distance to coast, and soil moisture. Country-wide maps of SIS and the corresponding pixel-level prediction errors at a spatial resolution of 12.5 m grid cells were produced for the two species. Independent validations show the site index maps are suitable for use in operational forest management planning in Sweden.