Regional Datasets Research Articles

Lithological mapping of charnockites using spectral mixture analysis

Charnockite, an orthopyroxene-bearing granitic rock, is the most dominant rock type of the Southern Granulite Terrain (SGT) of India. This study utilises remote sensing data to map the charnockite lithology in the Kodaikanal region within the SGT. Although remote sensing-based lithological mapping dominantly uses VNIR (Visible-Near Infrared) datasets, in this study, PRISMA (hyperspectral) datasets in the VNIR region and ASTER (multispectral) datasets in the TIR (Thermal Infrared) region were utilised in deriving the mineralogical composition and spectral characteristics of charnockites. In addition, image-derived endmember spectra from PRISMA and ASTER were validated using the laboratory-acquired spectra of charnockites. Linear mixing models (LMM) and the Hapke model were applied to quantitatively study the fractional abundance of charnockite rocks in the Kodaikanal region. Three least square methods, i.e., Fully Constrained Least Squares (FCLS), Non-Negative Least Squares (NNLS), and Unconstrained Least Squares (UCLS), were used to derive the fractional abundance of endmembers. Laboratory-derived charnockite spectra in the VNIR region exhibit spectral features characteristics of pyroxenes and hydrated minerals. In the TIR region, charnockite spectra exhibit characteristic vibrational absorption features of quartz and feldspar known as reststrahlen bands. Image analysis indicates that charnockite rocks exhibit spectral characteristics resembling the mafic rock type in the VNIR region and the felsic rock type in the TIR region. Therefore, the combined analysis of VNIR and TIR datasets allows the accurate delineation of charnockite lithology. However, vegetation in the study area introduces strong nonlinear spectral mixing effects, making it challenging to map the lithological units accurately. This study addresses this challenge using spectral mixture analysis to map charnockites in the Kodaikanal region. The fractional abundance study provides valuable insights into the lithological composition, demonstrating the effectiveness of spectral mixture analysis in mapping charnockites in the Kodaikanal region. Overall, the study demonstrates that the spectral mixture analysis significantly enhances the accuracy of lithology mapping, contributing to a more comprehensive understanding of the study area.

GPEP v1.0: the Geospatial Probabilistic Estimation Package to support Earth science applications

Abstract. Ensemble geophysical datasets are foundational for research to understand the Earth system in an uncertainty-aware context and to drive applications that require quantification of uncertainties, such as probabilistic hydro-meteorological estimation or prediction. Yet ensemble estimation is more challenging than single-value spatial interpolation, and open-access routines and tools are limited in this area, hindering the generation and application of ensemble geophysical datasets. A notable exception in the last decade has been the Gridded Meteorological Ensemble Tool (GMET), which is implemented in FORTRAN and has typically been configured for ensemble estimation of precipitation, mean air temperature, and daily temperature range, based on station observations. GMET has been used to generate a variety of local, regional, national, and global meteorological datasets, which in turn have driven multiple retrospective and real-time hydrological applications. Motivated by an interest in expanding GMET flexibility, application scope, and range of methods, we have developed the Python-based Geospatial Probabilistic Estimation Package (GPEP) that offers GMET functionality along with additional methodological and usability improvements, including variable independence and flexibility, an efficient alternative cross-validation strategy, internal parallelization, and the availability of the scikit-learn machine learning library for both local and global regression. This paper describes GPEP and illustrates some of its capabilities using several demonstration experiments, including the estimation of precipitation, temperature, and snow water equivalent ensemble analyses on various scales.

Improving Breast Tumor Multi-Classification from High-Resolution Histological Images with the Integration of Feature Space Data Augmentation

To support pathologists in breast tumor diagnosis, deep learning plays a crucial role in the development of histological whole slide image (WSI) classification methods. However, automatic classification is challenging due to the high-resolution data and the scarcity of representative training data. To tackle these limitations, we propose a deep learning-based breast tumor gigapixel histological image multi-classifier integrated with a high-resolution data augmentation model to process the entire slide by exploring its local and global information and generating its different synthetic versions. The key idea is to perform the classification and augmentation in feature latent space, reducing the computational cost while preserving the class label of the input. We adopt a deep learning-based multi-classification method and evaluate the contribution given by a conditional generative adversarial network-based data augmentation model on the classifier’s performance for three tumor classes in the BRIGHT Challenge dataset. The proposed method has allowed us to achieve an average F1 equal to 69.5, considering only the WSI dataset of the Challenge. The results are comparable to those obtained by the Challenge winning method (71.6), also trained on the annotated tumor region dataset of the Challenge.

A regional-scale approach to assessing non-residential building, transportation and cropland exposure in Central Asia

Abstract. Critical infrastructure has a paramount role in socio-economic development, and its disruption can have dramatic consequences for human communities, including cascading impacts. Assessing critical-infrastructure exposure to multiple hazard is therefore of utmost importance for disaster risk reduction purposes. However, past efforts in exposure assessment have predominantly concentrated on residential buildings, often overlooking the unique characteristics of critical infrastructure. Knowing the location, type and characteristics of critical infrastructure is particularly challenging due to the overall scarcity of data and difficulty of interacting with local stakeholders. We propose a method to assess exposure of selected critical infrastructure and demonstrate it for Central Asia, a region prone to multiple hazards (e.g., floods, earthquakes, landslides). We develop the first regionally consistent exposure database for selected critical infrastructure and asset types (namely, non-residential buildings, transportation and croplands), assembling the available global and regional datasets together with country-based information provided by local authorities and research groups, including reconstruction costs. The method addresses the main known challenges related to exposure assessment of critical infrastructure (i.e., data scarcity, difficulties in interacting with local stakeholders) by collecting national-scale data with the help of local research groups. The analysis also includes country-based reconstruction costs, supporting regional-scale disaster risk reduction strategies that include the financial aspect.

DNA Barcoding and Fertilization Strategies in Sideritis syriaca subsp. syriaca, a Local Endemic Plant of Crete with High Medicinal Value.

Herein, we applied DNA barcoding for the genetic characterization of Sideritis syriaca subsp. syriaca (Lamiaceae; threatened local Cretan endemic plant) using seven molecular markers of cpDNA. Five fertilization schemes were evaluated comparatively in a pilot cultivation in Crete. Conventional inorganic fertilizers (ChFs), integrated nutrient management (INM) fertilizers, and two biostimulants were utilized (foliar and soil application). Plant growth, leaf chlorophyll fluorescence, and color were assessed and leaf content of chlorophyll, key antioxidants (carotenoids, flavonoids, phenols), and nutrients were evaluated. Fertilization schemes induced distinct differences in leaf shape, altering quality characteristics. INM-foliar and ChF-soil application promoted yield, without affecting tissue water content or biomass partitioning to inflorescences. ChF-foliar application was the most stimulatory treatment when the primary target was enhanced antioxidant contents while INM-biostimulant was the least effective one. However, when the primary target is yield, INM, especially by foliar application, and ChF, by soil application, ought to be employed. New DNA sequence datasets for the plastid regions of petB/petD, rpoC1, psbK-psbI, and atpF/atpH were deposited in the GenBank for S. syriaca subsp. syriaca while the molecular markers rbcL, trnL/trnF, and psbA/trnH were compared to those of another 15 Sideritis species retrieved from the GenBank, constructing a phylogenetic tree to show their genetic relatedness.

Availability of Deceased Donors for Uterus Transplantation in the United States: Perception vs. Reality

Uterus transplantation (UTx) is a rapidly evolving treatment for uterine factor infertility. New centers offering this treatment must decide whether to utilize living donors, deceased donors, or both. Although limiting UTx to deceased donors eliminates the surgical risks for living donors, an adequate supply of suitable deceased uterus donors in the United States is an emerging concern. Previous studies describing the paucity of deceased uterus donors failed to consider key donor characteristics, potentially overestimating the available organ pool. To estimate the United States’ supply of deceased donor uteri; we extrapolated detailed clinical and demographic information from the regional donor datasets available from three organ procurement organizations to the national Organ Procurement and Transplantation Network donor pool. We estimate there are approximately 3700 possible and 400 optimal uterus donors annually in the United States. Given these projections and the number of women with uterine factor infertility in the U.S. who pursue parenthood through alternative strategies, we conclude that, as uterus transplant transitions from research to established clinical care, demand could quickly exceed the deceased donor supply. The liberalization of deceased donor selection criteria may be insufficient to address this imbalance; therefore, fulfilling the anticipated increased demand for uterus transplantation may require and justify greater use of living donors.

Threats of dams to the persistence of the world's freshwater fishes

AbstractHabitat fragmentation due to dams is a well‐known threat to freshwater fish. Yet, the global consequences of fragmentation for the viability of freshwater fish populations are unknown. Here, we provide the first global assessment of the threats of dams to the persistence of freshwater fish species. We developed a global macroecological relationship between freshwater fish range size and body size and used this relationship to assess whether isolated range fragments are too small to support a species. Our assessment includes 7369 freshwater fish species and considers the effects of 31,780 dams globally. Furthermore, we performed a more detailed analysis of the threats of dams in Brazil, the greater Mekong region and the United States, using complementary national and regional data sets. Globally, more than half of the species analysed face extirpation in a part of their geographic range, with an average potential range loss of 3.3% (95%‐range: 0%–31.8%) across all species analysed. For 74 fish species, occurring in Brazil, China, India, the Mekong basin, the United States, South‐Africa and the East Adriatic Coast, more than 50% of their range is potentially lost. This includes 18 species threatened with extinction across their whole range. Our complementary regional analysis revealed that the potentially lost range increases by a factor of 2–4 when considering both large and small dams compared with considering only large dams (≥15 m), highlighting the need to establish more comprehensive global dam inventories. Our novel approach and global analysis identifies species at risk of extirpation as well as geographic hotspots of extirpation threat by dams, which can aid in establishing more effective strategies for global hydropower development and barrier removal efforts to optimise the trade‐offs between biodiversity conservation and the socio‐economic benefits of dams.

Tool wear prediction through AI-assisted digital shadow using industrial edge device

Flank wear of drilling tools in manufacturing is among the main factors affecting product quality and productivity. In this study, an AI-assisted digital shadow was created for the instant prediction of the drilling flank tool wear. The drilling data were collected simultaneously using an industrial edge device and a rotary dynamometer. Feature engineering was conducted on the collected data from devices in time and frequency domains. A recurrent neural network (RNN) based on bidirectional long short-term memory (Bi-LSTM) and bidirectional gated recurrent unit (Bi-GRU) architectures was implemented on specified tool wear regions dataset. The digital shadow was created using the industrial edge device and the predictive AI model to minimize costs by reducing the need for expensive multi-sensors, manufacturing downtime, and tool underuse or overuse in a smart manufacturing system. The proposed model predicts with high accuracy and computational time efficiency and can be integrated into digital twin systems.

Australian Ocean surface waves dataset from SAR

AbstractIn this article, a regional ocean surface waves dataset from Sentinel‐1 A and B Synthetic Aperture Radar (SAR) satellites has been described. The ocean wave data have been extracted from the Sentinel‐1 level‐2 OCN (ocean) product as provided by the European Space Agency and downloadable for this region from the Copernicus Australasia regional data hub. The source OCN data have been produced by evolving versions of Sentinel‐1 Instrument Processing Facility (IPF). The structure of the source OCN NetCDF files changes over time and presents a challenge in performing long duration, time series analyses, including the examination of potential inconsistencies in OCN wave data, due to employment of different IPF versions over the duration of the satellite missions. Here, the input OCN wave data have been homogenized to a single, easily usable standard format after applying a quality assurance and control procedure that removes various inconsistencies in variables, coordinates, dimensions and land flag, and through the addition of new auxiliary variables. The new format has the desirable properties of being compact in size, consistent in structure, and scalable in temporal and spatial coverage. It is also convenient to use and offers opportunities to perform fast, multi‐year regional processing and analysis for calibration and validation studies and scientific applications. No re‐processing of Sentinel‐1 level‐1 data has been carried out in this work.

Deep evolutionary fusion neural network: a new prediction standard for infectious disease incidence rates

BackgroundPreviously, many methods have been used to predict the incidence trends of infectious diseases. There are numerous methods for predicting the incidence trends of infectious diseases, and they have exhibited varying degrees of success. However, there are a lack of prediction benchmarks that integrate linear and nonlinear methods and effectively use internet data. The aim of this paper is to develop a prediction model of the incidence rate of infectious diseases that integrates multiple methods and multisource data, realizing ground-breaking research.ResultsThe infectious disease dataset is from an official release and includes four national and three regional datasets. The Baidu index platform provides internet data. We choose a single model (seasonal autoregressive integrated moving average (SARIMA), nonlinear autoregressive neural network (NAR), and long short-term memory (LSTM)) and a deep evolutionary fusion neural network (DEFNN). The DEFNN is built using the idea of neural evolution and fusion, and the DEFNN + is built using multisource data. We compare the model accuracy on reference group data and validate the model generalizability on external data. (1) The loss of SA-LSTM in the reference group dataset is 0.4919, which is significantly better than that of other single models. (2) The loss values of SA-LSTM on the national and regional external datasets are 0.9666, 1.2437, 0.2472, 0.7239, 1.4026, and 0.6868. (3) When multisource indices are added to the national dataset, the loss of the DEFNN + increases to 0.4212, 0.8218, 1.0331, and 0.8575.ConclusionsWe propose an SA-LSTM optimization model with good accuracy and generalizability based on the concept of multiple methods and multiple data fusion. DEFNN enriches and supplements infectious disease prediction methodologies, can serve as a new benchmark for future infectious disease predictions and provides a reference for the prediction of the incidence rates of various infectious diseases.

Assessing Healthy Effects between Local Level Farmer’s Markets and Community-Supported Agriculture and Physical Well-Being at the State Level

Much change has occurred in the recent past around food systems at the local level. As interest continues to rise, we were curious to examine connections of local food systems and well-being in the context of the places they operate at the state level. Health is a central component of food well-being. In this article, we focus on physical well-being. We explore the connection between food systems, as represented by farmers markets (FMs), community-supported agriculture (CSAs), and physical well-being using state level data. To investigate these relationships, we conducted an ordinal logistic regression model analyzing 2016 data from the United States Department of Agriculture for the 50 states. The results of this study suggest that there is an inconclusive relationship at the state level. While local FMs and CSAs may increase physical well-being at the state level, we did not find this in our study. This finding adds to the body of literature where the majority of the studies in this area are case studies, ethnographies, and smaller studies using some regional data and small samples for local data. We believe that this conceptual work will allow for future empirical studies to build upon this study to conduct further empirical work examining regional state and local datasets for more conclusive findings.

Industrial robots, and information and communication technology: the employment effects in EU labour markets

ABSTRACT This paper explores the effects of industrial robots and information and communication technology (ICT) on regional employment in European Union countries. The empirical analysis relies on a harmonised comprehensive regional dataset that combines business statistics and national and regional accounts data. This rich dataset enables us to provide detailed insights into the employment effects of automation and computerisation in EU regions for the period 2001–16. The results suggest relatively weak effects on regional total employment dynamics. However, industrial robots show negative employment effects in local manufacturing industries and positive employment effects in local non-manufacturing industries. While the negative effect is concentrated in particular local manufacturing industries, the positive effect has operated in local service industries. Information technology investments show positive employment effects in local manufacturing industries and some individual local service industries, while communication technology investments are shown to be irrelevant for employment dynamics. In contrast, software and database investments have had a predominantly negative association with local employment.

Assessment of ocean color atmospheric correction methods and development of a regional ocean color operational dataset for the Baltic Sea based on Sentinel-3 OLCI

The Baltic Sea is characterized by large gradients in salinity, high concentrations of colored dissolved organic matter, and a phytoplankton phenology with two seasonal blooms. Satellite retrievals of chlorophyll-a concentration (chl-a) are hindered by the optical complexity of this basin and the reduced performance of the atmospheric correction in its highly absorbing waters. Within the development of a regional ocean color operational processing chain for the Baltic Sea based on Sentinel-3 Ocean and Land Colour Instrument (OLCI) full-resolution data, the performance of four atmospheric correction processors for the retrieval of remote-sensing reflectance (Rrs) was analyzed. Assessments based on three Aerosol Robotic Network-Ocean Color (AERONET-OC) sites and shipborne hyperspectral radiometers show that POLYMER was the best-performing processor in the visible spectral range, also providing a better spatial coverage compared with the other processors. Hence, OLCI Rrs spectra retrieved with POLYMER were chosen as input for a bio-optical ensemble scheme that computes chl-a as a weighted sum of different regional multilayer perceptron neural nets. This study also evaluated the operational Rrs and chl-a datasets for the Baltic Sea based on OC-CCI v.6. The chl-a retrievals based on OC-CCI v.6 and OLCI Rrs, assessed against in-situ chl-a measurements, yielded similar results (OC-CCI v.6: R2 = 0.11, bias = −0.22; OLCI: R2 = 0.16, bias = −0.03) using a common set of match-ups for the same period. Finally, an overall good agreement was found between chl-a retrievals from OLCI and OC-CCI v.6 although differences between Rrs were amplified in terms of chl-a estimates.

Evaluation of precipitation across the contiguous United States, Alaska, and Puerto Rico in multi-decadal convection-permitting simulations

This study is an early effort to generate a multi-decadal convection-permitting regional climate dataset that covers nearly the entire North American continent. We assessed a 20 year dynamically downscaled regional climate simulation at a 4 km spatial resolution with explicit convection across the contiguous United States (CONUS), Alaska, and Puerto Rico. Specifically, we evaluated the model’s performance in representing mean, 95th percentile, and extreme precipitation across regions. Our findings indicate that when compared with ERA5 reanalysis, the forcing data, convection-permitting simulation improves representations of seasonal, 95th percentile, and extreme precipitation over a large portion of the CONUS, Alaska, and Puerto Rico, particularly in areas where precipitation is heaviest. The simulation adds value over its forcing data (ERA5) in up to 53% of all grid cells in the CONUS, 68.8% in Alaska, and 84.0% in Puerto Rico. It is important to note that, however, despite improvements, model errors in Puerto Rico remain large. Similar improvements are observed in extreme indices, including consecutive dry days, maximum 5 days precipitation, and extreme precipitation. Analysis of the diurnal cycle of mean hourly precipitation suggests that representations of convective processes—including onset, dissipation, suppression, downstream propagation, and local circulation—improved overall.

Mineral exploration and regional surface geochemical datasets: An anomaly detection and k-means clustering exercise applied on laterite in Western Australia

A comprehensive geochemical survey was conducted in the western Yilgarn Craton, Western Australia, in 2007, collecting 3142 surface samples of regolith. Our study used this data to target potential sites for undiscovered buried or concealed Cu-Zn-Pb, and NiCu deposits. The core approach used the singularity mapping technique for detecting anomalies at the local scale. This work proposes a procedure to create a composite multi-element singularity map by linearly combining individual element singularity maps, using element-to-element correlation coefficients as weights for the linear combination process. Furthermore, the k-means clustering algorithm was applied to combinations of sub-sets of data and singularity values. Expert validation indicated that the k-means clustering approach yielded the best results when using 4 or 5 clusters, separating the distinct sites of the MINEDEX database. In either case, the incorporation of the singularity values provided the most accurate outcomes, with a dominant cluster correctly classifying up to 60 to 80 % of identified Cu and Ni deposits and mines, respectively.Based on these results and on the range of computed singularity values, simple rules were established to identify sampled data points satisfying the following criteria: (i) meeting the defined threshold singularity value and belonging to the k-means cluster that include the mines and (ii) not being in the neighbourhood of any known mineralization site from the MINEDEX database. These locations thus represent potential mineralization sites that warrant further investigation and exploration follow-up. The outcomes of this study strongly support the efficiency of anomaly detection and k-means clustering method applied on a regional surface geochemical dataset for mineral exploration to detect and target mineral systems.

A global dataset on phosphorus in agricultural soils

Numerous drivers such as farming practices, erosion, land-use change, and soil biogeochemical background, determine the global spatial distribution of phosphorus (P) in agricultural soils. Here, we revised an approach published earlier (called here GPASOIL-v0), in which several global datasets describing these drivers were combined with a process model for soil P dynamics to reconstruct the past and current distribution of P in cropland and grassland soils. The objective of the present update, called GPASOIL-v1, is to incorporate recent advances in process understanding about soil inorganic P dynamics, in datasets to describe the different drivers, and in regional soil P measurements for benchmarking. We trace the impact of the update on the reconstructed soil P. After the update we estimate a global averaged inorganic labile P of 187 kgP ha−1 for cropland and 91 kgP ha−1 for grassland in 2018 for the top 0–0.3 m soil layer, but these values are sensitive to the mineralization rates chosen for the organic P pools. Uncertainty in the driver estimates lead to coefficients of variation of 0.22 and 0.54 for cropland and grassland, respectively. This work makes the methods for simulating the agricultural soil P maps more transparent and reproducible than previous estimates, and increases the confidence in the new estimates, while the evaluation against regional dataset still suggests rooms for further improvement.

Analysis of the inflection points of height-diameter models

The inflection point is an important feature of sigmoidal height-diameter (H-D) models. It is often cited as one of the properties favoring sigmoidal model forms. However, there are very few studies analyzing the inflection points of H-D models. The goals of this study were to theoretically and empirically examine the behaviors of inflection points of six common H-D models with a regional dataset. The six models were the Wykoff (WYK), Schumacher (SCH), Curtis (CUR), Hossfeld IV (HOS), von Bertalanffy-Richards (VBR), and Gompertz (GPZ) models. The models were first fitted in their base forms with tree species as random effects and were then expanded to include functional traits and spatial distribution. The distributions of the estimated inflection points were similar between the two-parameter models WYK, SCH, and CUR, but were different between the three-parameter models HOS, VBR, and GPZ. GPZ produced some of the largest inflection points. HOS and VBR produced concave H-D curves without inflection points for 12.7% and 39.7% of the tree species. Evergreen species or decreasing shade tolerance resulted in larger inflection points. The trends in the estimated inflection points of HOS and VBR were entirely opposite across the landscape. Furthermore, HOS could produce concave H-D curves for portions of the landscape. Based on the studied behaviors, the choice between two-parameter models may not matter. We recommend comparing several three-parameter model forms for consistency in estimated inflection points before deciding on one. Believing sigmoidal models to have inflection points does not necessarily mean that they will produce fitted curves with one. Our study highlights the need to integrate analysis of inflection points into modeling H-D relationships.

Health Inequality Data Repository

The Health Inequality Data Repository from the World Health Organization (WHO) is a free, open-access selection of publicly available disaggregated data about health and indicators of health. It includes regional and continental data sets from 1950 to 2022, covering all world regions, from 16 sources on more than 2,000 indicators. As of November 2023, the available data set categories are Sustainable Development Goals; WHO Global Health Observatory; COVID-19; Reproductive, Maternal, and Child Health; Immunization; HIV, Tuberculosis, and Malaria; Adult Health; Health Care; Burden of Disease; Disability; Migration; Environmental Health; WHO General Programme of Work; and Beyond the Health Sector (Health Inequality Monitor 2023). The sources of the data sets include publicly available data from the WHO and numerous United Nations (UN) agencies, the Organisation for Economic Co-operation and Development (OECD), the European Union, the US Agency for International Development (USAID), and more. In addition to the data sets, two software options are available. The Health Equity Assessment Toolkit (HEAT) allows users to view the data in various chart and graph formats, and HEAT Plus allows users to upload their own disaggregated data in addition to the available data sets. The Health Inequality Data Repository and its supporting software tools are housed within the Health Inequality Monitor section of the WHO website, and it has numerous educational options related to its resources and goals.

Spatial and Temporal Exhumation of the Northeastern China: Insights From Low Temperature Thermochronology

ABSTRACTThe northeastern (NE) China has undergone a long‐term tectonic history involving subduction–accretion of the Siberian Craton, North China Craton, and Western Pacific Plate, providing an ideal target for understanding the interactions between cratons and the Pacific oceanic lithosphere. However, the Mesozoic‐Cenozoic geological and topographic evolutions of the NE China remain unclear. Here, we present new apatite fission track data from Precambrian metamorphic rocks in the Lesser Hinggan Mountains. Apatite fission track dating yielded a group of central ages ranging from 130 ± 18 to 151 ± 11 Ma with mean track lengths varying between 13.67 ± 0.13 and 14.49 ± 0.08 μm. Thermal modeling results revealed a rapid cooling event at ca. 165–140 Ma, which was related to compressional deformations driven by shallow subduction of the Paleo‐Pacific Ocean Plate beneath the northeastern China. Furthermore, regional low temperature thermochronological datasets in NE China show additional three rapid exhumation stages, including the Early Jurassic (ca. 200–180 Ma), Early Cretaceous (ca. 120–100 Ma), and Late Cretaceous (ca. 80–60 Ma). They were associated with closure of the Mudanjiang Ocean, slab rollback of the Paleo‐Pacific Plate, and collision between the Okhotomorsk Block and southeastern Asia. As the Cenozoic, slow exhumation has emerged throughout the northeastern China.

Multi-step power forecasting for regional photovoltaic plants based on ITDE-GAT model

With the integration of high proportion of distributed photovoltaic(PV), high-accuracy regional PV power forecasting technology can enhance the regional coordinated scheduling capability of the new power system. This paper proposes a regional PV power forecasting model based on an improved time-series dense encoder and graph attention network (ITDE-GAT), which takes into account the spatio-temporal correlations among the regional PV plants. Firstly, an improved complete ensemble empirical mode decomposition with adaptive noise(ICEEMDAN) is used to extract the clear-sky and fluctuation components from PV data. Secondly, the combined ITDE-GAT is applied to perform the regional PV power forecasting. Considering the static information, an improved dense encoder network (ITDE) is constructed to extract the temporal and spatial relationships of regional PV. Graph attention network (GAT) is then utilized to explore the spatial correlations among the regional PV. Finally, case study from two actual PV region datasets shows that the proposed model achieves higher forecasting accuracy and exhibits stronger generalization capabilities. The results demonstrate that compared to various advanced deep learning methods, the R2 evaluation metric of the approach proposed in this paper demonstrates, respectively, maximum improvements of 3.4 %, 6.5 %, and 7.8 % for the 1 h, 3 h, and 6 h ahead predictions.