Temperature Data Research Articles

Cross-continental comparison of plant reproductive phenology shows high intraspecific variation in temperature sensitivity

Abstract The success of plant species under climate change will be determined, in part, by their phenological responses to temperature. Despite the growing need to forecast such outcomes across entire species ranges, it remains unclear how phenological sensitivity to temperature might vary across individuals of the same species. In this study, we harnessed community science data to document intraspecific patterns in phenological temperature sensitivity across the multicontinental range of six herbaceous plant species. Using linear models, we correlated georeferenced temperature data with 23,220 plant phenological records from iNaturalist to generate spatially-explicit estimates of phenological temperature sensitivity across the shared range of species. We additionally evaluated the geographic association between local historic climate conditions (i.e. mean annual temperature and interannual variability in temperature) and the temperature sensitivity of plants. We found that plant temperature sensitivity varied substantially at both the interspecific and intraspecific level, demonstrating that phenological responses to climate change have the potential to vary both within and among species. Additionally, we provide evidence for a strong geographic association between plant temperature sensitivity and local historic climate conditions. Plants were more sensitive to temperature in hotter climates (i.e., regions with high mean annual temperature), but only in regions with high interannual temperature variability. In regions with low interannual temperature variability, plants displayed universally weak sensitivity to temperature, regardless of baseline annual temperature. This evidence suggests that pheno-climatic forecasts may be improved by accounting for intraspecific variation in phenological temperature sensitivity. Broad climatic factors such as mean annual temperature and interannual temperature variability likely serve as useful predictors for estimating temperature sensitivity across species’ ranges.

Performance Evaluation of CMIP6 Climate Model Projections for Precipitation and Temperature in the Upper Blue Nile Basin, Ethiopia

The projection and identification of historical and future changes in climatic systems is crucial. This study aims to assess the performance of CMIP6 climate models and projections of precipitation and temperature variables over the Upper Blue Nile Basin (UBNB), Northwestern Ethiopia. The bias in the CMIP6 model data was adjusted using data from meteorological stations. Additionally, this study uses daily CMIP6 precipitation and temperature data under SSP1-2.6, SSP2-4.5, and SSP5-8.5 scenarios for the near (2015–2044), mid (2045–2074), and far (2075–2100) periods. Power transformation and distribution mapping bias correction techniques were used to adjust biases in precipitation and temperature data from seven CMIP6 models. To validate the model data against observed data, statistical evaluation techniques were employed. Mann–Kendall (MK) and Sen’s slope estimator were also performed to identify trends and magnitudes of variations in rainfall and temperature, respectively. The performance evaluation revealed that the INM-CM5-0 and INM-CM4-8 models performed best for precipitation and temperature, respectively. The precipitation projections in all agro-climatic zones under SSP1-2.6, SSP2-4.5, and SSP5-8.5 scenarios show a significant (p < 0.01) positive trend. The mean annual maximum temperature over UBNB is estimated to increase by 1.8 °C, 2.1 °C, and 2.8 °C under SSP1-2.6, SSP2-4.5, and SSP5-8.5 between 2015 and 2100, respectively. Similarly, the mean annually minimum temperature is estimated to increase by 1.5 °C, 2.1 °C, and 3.1 °C under SSP1-2.6, SSP2-4.5, and SSP5-8.5, respectively. These significant changes in climate variables are anticipated to alter the incidence and severity of extremes. Hence, communities should adopt various adaptation practices to mitigate the effects of rising temperatures.

Multivariate stochastic generation of meteorological data for building simulation through interdependent meteorological processes

In recent years, the uncertainty of weather conditions and the impact of future climate change on building energy assessment has received increasing attention. As an important part of these studies, several types of methods for generating stochastic meteorological data have also been developed. Since solar radiation drops to zero at night, unlike the continuous 24-hour data for elements such as temperature and humidity, this has posed challenges for previous research to fully account for the simultaneity among multiple elements. Therefore, this study proposes a framework for meteorological data generation: First, perform multivariate time series modeling of meteorological data of air temperature, solar radiation and absolute humidity at 12:00 of each day of a typical year based on the S-vine copula method and simulating daily series data at 12:00 for 365 days. Then, based on the probability of change of each element evaluated from the historical meteorological observation data, the daily series data at 12:00 were expanded to 24 h, after which the yearly stochastic weather data were obtained. The analysis of 30 years of stochastic data generated by this method, compared with the original data, reveals that air temperature and solar radiation closely match the original distribution characteristics, except for a minor deviation in the absolute humidity’s kurtosis. Furthermore, the comparison of thermal load distributions for office buildings shows that the original data curve falls within the range of the generated data. This suggests that the generated data includes more information about uncertainty but still keeps the original data’s characteristics.

Optimizing wheat supplementary irrigation: Integrating soil stress and crop water stress index for smart scheduling

A two-year field experiment was conducted to integrate soil moisture stress with the Crop Water Stress Index (CWSI) for optimizing irrigation in winter wheat (Triticum aestivum L.) under varying irrigation regimes. The study took place at the Water Technology Centre (WTC-02) of ICAR-IARI, New Delhi, where the climate shows a blend of monsoon-influenced humid subtropical and semi-arid conditions. Using a randomized block design (RBD), five irrigation treatments were applied: full irrigation and deficit irrigation (DI) at 15 %, 30 %, 45 %, and 60 % levels. Canopy and ambient air temperature data, along with vapor pressure deficit (VPD), were recorded using a developed integrated sensing device to empirically determine the lower baseline equations and upper threshold for CWSI computation at pre-heading and post-heading stages. The slope (m), intercept (c) of the lower baseline equation, and upper threshold (UL) for pre-heading and post-heading were found: m: −1.94, c: −1.33, UL: 1.92°C and m: −1.30, c: −2.37, UL: 2.0°C, respectively. Results showed that increasing water deficit levels led to significant reductions in grain yield, biomass production, and harvest index. A strong negative correlation (R² = 0.95 and 0.93) between mean seasonal CWSI and yield attributes highlighted the utility of CWSI in yield prediction under varying irrigation regimes. It is recommended to schedule irrigation based on the CWSI approach when CWSI ≥0.35 for optimum wheat yields. Integrating CWSI with soil moisture stress provides valuable real-time insights into crop water status, enabling more precise and smart irrigation scheduling.

Modelling Marine Heatwaves Impact on Shallow and Upper Mesophotic Tropical Coral Reefs

Abstract Coral reefs ecosystems, often compared to rain forests for their high biodiversity, are threatened by ocean warming causing coral bleaching when the symbiotic relationship between dinoflagellates and corals breaks under high ocean temperatures. Thermal stress from marine heatwaves occur both at the surface and subsurface with subsurface marine heatwaves lasting longer with potentially higher cumulative intensities. However, global coral bleaching models generally ignore the differences in thermal stress between surface and sea-bed levels. Here, we define marine heatwaves at sea-bed level to model coral bleaching with daily resolution from May 6th 1993 to December 31st 2023, for 9944 tropical coral reefs between 0 and 50m depths. We show that deeper reefs experience on average higher thermal stress and bleaching compared to surface reefs. Using surface temperature data to model bleaching for deeper corals underestimates bleaching intensities by an average of 6 ± 9% compared to the subsurface calibrated model. Our study is a starting point for more accurate coral bleaching modelling, providing additional evidence to reshape our perception of deeper coral reefs as potential refugees from climate change. 

Innovation for recycling of organic matter through composter with automatic and sustainable temperature recording accessed via Bluetooth/mobile app.

Compost reactors, commonly used in experiments, industrial assays, and home residue treatment systems, have the potential to facilitate composting. Challenges persist in the realm of small-scale composting, encompassing facets such as temperature monitoring, homogenization of the compost mass, management of moisture with the control of leachate generation, and integration with a renewable energy source. This study assesses a pioneering composter prototype endowed with essential features to ensure a pragmatic and secure composting process. This includes the facilitation of remote access to temperature data via Bluetooth and a mobile application. Across successive trials, the scrutinized composter prototype consistently yielded reproducible outcomes, exhibiting a coefficient of variation below 25% for the majority of appraised parameters. In comparison to a conventional reactor, the decomposing residue mixture within the examined prototype manifested elevated temperatures (p < 0.05). Moreover, the tested prototype demonstrated C/N ratio lower than 20/1 within 45days, a higher final nitrogen concentration, and enhanced germination of seeds that served as phytotoxicity bioindicators. Notably, the prototype needed 46.6% less space, offering improved leachate control, three times faster turning time, temperature monitoring, and reduced fly attraction.

Dew Point Characteristics at Synoptic Stations in Northern Benin

The dew point temperature is a very important parameter for hydro and agro-climatological research. This work studied the temporal variability of dew point temperature in Northern Benin. The dew point data comes from three synoptic stations in Northern Benin and covers the period from January 1980 to February 2019. After, correction and homogenization of the data, statistical methods are employed to analyze its structure, distribution, and temporal variability. The average of the dataset is 16.607°C with a standard deviation of 6.871 and a coefficient of variation of 0.414. The distribution is negatively skewed without extreme values, with an interquartile range of 10.284. The analysis by days of month indicates an irregular change in values at the beginning of month. The days of week show a minimum at the beginning and a maximum at the end of week. The analysis of dew point temperature data by months of the year shows a bell-shaped distribution with a plateau covering the months from May to September. The average dew point temperature has been increasing over the years. The data also describe a non-stationary periodic series approved by statistical tests. The absence of an inflection point in the data and in the trend means that the distribution is evolving and cyclical but not regular.

Compound Hot and Dry Events in Argentina and Their Connection to El Niño‐Southern Oscillation

ABSTRACTThis work studies the simultaneous and sequential occurrence of hot and dry months in the summer season in Argentina, north of 40°S, based on three different databases: meteorological stations, a gridded observational dataset and a reanalysis product. The influence of the El Niño‐Southern Oscillation (ENSO) over the occurrence of these compound events is specially analysed using a logistic regression model. Monthly maximum temperature and precipitation data are used for the period 1979–2022 in four sub‐regions of Argentina: Northwestern Argentina (NOA), Northeastern Argentina (NEA), Cuyo (central‐western Argentina) and the Pampas (central‐eastern Argentina). Simultaneous hot and dry months and hot months preceded by dry months are the most frequent compound events. The highest frequencies are found in the centre part of the study region and NEA for simultaneous compound events, and in NOA and the Pampas region for sequential ones. In general terms, all datasets show a good representation of the spatio‐temporal variability of hot and dry months. The insights of the influence of ENSO on compound events revealed that La Niña enhances the occurrences of hot and dry months throughout the study region, with the exception of NOA, where El Niño conditions promote the occurrence of these events. Based on logistic regression models, we successfully quantify the relationship between ENSO and hot and dry months and demonstrate that ENSO plays a significant role as a driver of compound hot and dry events in the central region, Cuyo, NEA and a portion of the Pampas. This research contributes to the understanding of compound events in Argentina and how they are influenced by major drivers of climate variability providing useful information for the development of a predictive system for such events.

Development of an implantable sensor system for in vivo strain, temperature, and pH monitoring: comparative evaluation of titanium and resorbable magnesium plates

Biodegradable magnesium is a highly desired material for fracture fixation implants because of its good mechanical properties and ability to completely dissolve in the body over time, eliminating the need for a secondary surgery to remove the implant. Despite extensive research on these materials, there remains a dearth of information regarding critical factors that affect implant performance in clinical applications, such as the in vivo pH and mechanical loading conditions. We developed a measurement system with implantable strain, temperature, pH and motion sensors to characterize magnesium and titanium plates, fixating bilateral zygomatic arch osteotomies in three Swiss alpine sheep for eight weeks. pH 1–2 mm above titanium plates was 6.6 ± 0.4, while for magnesium plates it was slightly elevated to 7.4 ± 0.8. Strains on magnesium plates were higher than on titanium plates, possibly due to the lower Young's modulus of magnesium. One magnesium plate experienced excessive loading, which led to plate failure within 31 h. This is, to our knowledge, the first in vivo strain, temperature, and pH data recorded for magnesium implants used for fracture fixation. These results provide insight into magnesium degradation and its influence on the in vivo environment, and may help to improve material and implant design for future clinical applications.

Identification of position, size and dynamic heat power of chips embedded in a real electronic board

Temperature is a critical factor in the lifespan of electronic systems. Therefore, it is essential to control all thermal parameters. In certain applications, such as commercial off-the-shelf (COTS) components, manufacturers may not have precise knowledge of the composition of electronic boards. As a result, it is important for them to define the positions of heat-dissipating elements, and the power dissipated to better predict the final thermal behavior of the system. Some electronic boards today even have components integrated into their volume. Since components can now be located not only on the surface but also within the volume of electronic boards, manufacturers need to define volume-based methods for detecting these heat sources. In this article, we address the problem of detecting sources within a volume using surface temperature data. To achieve this goal, we develop a representative numerical model of the system that can be used for an inverse source identification procedure. In a first time, we present a numerical approach that demonstrates the feasibility of the identification procedure on a board containing several electronic chips. Secondly, the volume source identification method is applied to a real case, using surfaces temperatures measured by infrared cameras. The experimental results obtained are consistent and show that this technique is suitable for detecting volumetric sources within an electronic board.

Real-time seepage health monitoring using spatial autocorrelation of distributed temperature data from fiber optic sensor

Real-time seepage health monitoring using spatial autocorrelation of distributed temperature data from fiber optic sensor

Optimization of Productivity of Fodder Crops with Green Conveyor System in the Context of Climate Instability in the North Kazakhstan Region

One of the main challenges in modern animal husbandry in North Kazakhstan is ensuring an uninterrupted supply of sufficient fodder crops. This research, conducted from 2019 to 2023, aimed to develop strategies for cultivating environmentally sustainable fodder crops capable of providing a stable fodder crop base under the changing climatic conditions of the North Kazakhstan region. The studies included analysis of air temperature and precipitation data as well as monitoring of fodder grass mixtures within a green fodder conveyor system. Different sowing dates for fodder crops and mixtures were selected for the development of the conveyor system. The range of experimental variants included fodder crops and their mixtures from various botanical families. The experiment involved both perennial (alfalfa and festulolium) and annual (corn, pea, sunflower, Sudan grass, oats, and rapeseed) crops. The highest green mass yields were achieved by the following variants: fodder crops of corn + pea—74.40 c/ha; mixtures of annual legume–grass crops in the pea + oats variant of the first sowing date—43.64 c/ha; Sudan grass + pea—45.72 c/ha; mixtures of perennial grasses in the second utilization term of alfalfa + festulolium—64.9 c/ha; and rapeseed sown at the first sowing date—46.61 c/ha. In terms of crude and digestible protein content, the best among the annual grass variants was the mixture of Sudan grass and pea (crude protein—33.59 g/kg, digestible protein—24.5 g/kg), and the best among the perennials was the variant of the first utilization term (crude protein—50.42 g/kg, digestible protein—38.2 g/kg). Regarding metabolizable energy content, the annual crop variant of corn + pea had a yield of 1.92 MJ/kg, and in the perennial variant, the mixture of alfalfa and festulolium in the first utilization term had a yield of 2.68 MJ/kg. Such an approach to creating green fodder conveyors can be crucial for developing effective strategies for adapting agriculture to climate change, including the selection of promising fodder crops and optimization of their placement. The results obtained can contribute to enhancing the productivity and sustainability of agricultural production in the North Kazakhstan region.

Enhancing artificial permafrost table predictions using integrated climate and ground temperature data: A case study from the Qinghai-Xizang highway

The complexities of permafrost changes, driven by climate warming and engineering activities, coupled with challenges in data acquisition, make it crucial and challenging to accurately predict the artificial permafrost table, particularly for subgrades in high-temperature unstable permafrost regions. To address this, this study developed a hybrid machine learning model (RF-LSTM-XGBoost) for permafrost table prediction. By analyzing climate change and ground temperature data from various positions and depths along the subgrade in the Tuotuo River section of the Qinghai-Xizang Highway, the Spearman correlation coefficient method was initially used to determine the important influencing factors. Random Forest (RF), Long Short-Term Memory Neural Network (LSTM), and Extreme Gradient Boosting (XGBoost) models were used to predict the artificial permafrost table, and grid search and cross-validation methods were employed to optimize the hyperparameters of each model. A linear weighted combination method based on the minimum cumulative absolute error was utilized to merge the models, and its performance was compared with the individual RF, LSTM, and XGBoost models. Subsequently, the feature importance of the variables in the machine learning model was analyzed. The results indicated a strong correlation between artificial permafrost table changes and factors such as daily average atmospheric temperature, subgrade surface ground temperature, and subgrade surface ground heat flux during the freezing-thawing cycle. The combined model highlighted daily atmospheric temperature as the most influential predictor, followed by ground heat flux, with the surface ground temperature being less impactful. The combined model demonstrated improved predictive accuracy, with MSE, MAPE, RMSE, MAE, and R2 values of 0.003, 0.052, 0.0085, 0.029, and 0.989, respectively, surpassing those of individual models. This model offers a rapid, accurate, and reliable approach for permafrost table prediction, advancing subgrade stability research in challenging permafrost environments.

SARIMA Model-Based Maximum Temperature Forecasting in Bangladesh: A Data-Driven Evaluation from 1981 to 2024

Bangladesh is a tropical nation where there are notable seasonal temperature changes. The Seasonal Autoregressive Integrated Moving Average (SARIMA) model is used in this study to forecast Bangladesh&amp;apos;s maximum temperature from 2023 to 2042. The objective is to assess how rising temperatures can affect public health, energy consumption, and agriculture. Autocorrelation and partial autocorrelation analysis will be used to improve the model. Analysis was done using historical maximum temperature data spanning from 1981 to 2022. Forecasts were produced using the SARIMA model, whose parameters were chosen in accordance with plots of the autocorrelation function (ACF) and partial autocorrelation function (PACF). The model SARIMA (1,1,2)(0,0,1) is selected based on AIC. In order to account for forecast uncertainty, forecasts were created for the years 2023–2042. 95% prediction ranges were then calculated. Bangladesh&amp;apos;s maximum temperatures are predicted by the SARIMA model to rise gradually, from roughly 33.75°C in 2023 to 34.17°C in 2042. With some degree of uncertainty, the 95% prediction intervals show a steady increasing trend between 33.53°C and 34.51°C. The anticipated increase in the highest temperatures has major consequences for Bangladesh. These results highlight how crucial it is to create adaptation plans and laws in order to lessen the effects of warming temperatures and increase resilience.

Assessing rainfall threshold for shallow landslides triggering: a case study in the Alpes Maritimes region, France

AbstractIn this work, we use a statistical approach for modeling shallow landslide rainfall thresholds (Caine 1980) with a case study for the Alpes-Maritimes region (France). Cumulated rainfall / duration (ED) thresholds are obtained with the CTRL-T algorithm (Melillo and al. 2018) for different non-exceedance probabilities from a landslide and two climatic datasets. This tool allows to automatically define rainfall events that might trigger landslides, ensuring robustness and objectivity in this process. The first climate dataset stores high resolution gridded rainfall data (1km resolution, hourly), which provides rainfall data with high temporal and spatial accuracy. This dataset, coming from radar data, is calibrated with rainfall gauges, ensuring a higher accuracy of the rainfall measurements. It provides the rainfall records directly used in the threshold construction The second dataset contains lower resolution gridded rainfall, snow, temperature, and evapotranspiration data (8km resolution, daily); it enables to assess the region’s climate through parameters imported in CTRL-T. The thresholds are then validated using a method designed by Gariano and et al. (2015). Several improvements are made to the initial method. First, evapotranspiration values approximated in the process are replaced by values from the second climate dataset, the result accounting best for the regional climate. Then, computing duration values used for isolating events and sub-events for each mesh point allows to consider the heterogeneity of the Alpes-Maritimes climate. Rainfall thresholds are eventually obtained, successively from a set of probable conditions (MRC) and a set of highly probable conditions (MPRC). The validation process strengthens the analysis as well as enables to identify best performing thresholds. This work represents novel scientific progress towards landslide reliable warning systems by (a) making a case study of empirical rainfall thresholds for Alpes-Maritimes, (b) using high-resolution rainfall data and (c) adapting the method to climatically heterogeneous zones.

Evaluating the spatiotemporal patterns of drought characteristics in a semi‐arid region of Limpopo Province, South Africa

Drought is a complex phenomenon resulting from below-average rainfall and is characterized by frequency, duration, and severity, occurring at a regional scale with dire consequences, especially in semiarid environments. This study used the Reconnaissance Drought Index (RDI) to assess drought severity in two district municipalities in Limpopo Province. Rainfall and air temperature data from 12 stations covering 1970–2020 were obtained from the Agricultural Research Council. The calculation of RDI relies on the monthly accumulation ratio of total rainfall to potential evapotranspiration (PET). For this study, PET was estimated using the Hargreaves and Samani temperature-based approach. The RDI results showed a high spatial–temporal variation in drought characteristics over the study area. All stations experienced extreme drought conditions in different years, with the maximum drought severity (-3.40) occurring from 2002–2003 in the western parts of the study area, indicating extreme drought. Furthermore, the results revealed continuous drought conditions over various periods, including severe droughts between 1995 and 1998 and between 2014 and 2016, with the severity varying between mild and moderate drought conditions. The results reveal notable but nonuniform drought patterns as the climate evolves, with potential implications for water availability and livelihoods. The study's findings underscore the significance of adopting multidimensional approaches to drought assessment that encompass meteorological and hydrological factors to inform strategies for adaptive water management and policy formulation in the face of a changing climate.

Modeling Temperature-Dependent Photoluminescence Dynamics of Colloidal CdS Quantum Dots Using Long Short-Term Memory (LSTM) Networks

This study addresses the challenge of modeling temperature-dependent photoluminescence (PL) in CdS colloidal quantum dots (QD), where PL properties fluctuate with temperature, complicating traditional modeling approaches. The objective is to develop a predictive model capable of accurately capturing these variations using Long Short-Term Memory (LSTM) networks, which are well suited for managing temporal dependencies in time-series data. The methodology involved training the LSTM model on experimental time-series data of PL intensity and temperature. Through numerical simulation, the model’s performance was assessed. Results demonstrated that the LSTM-based model effectively predicted PL trends under different temperature conditions. This approach could be applied in optoelectronics and quantum dot-based sensors for enhanced forecasting capabilities.

Design and implementation of an IoT-based monitoring system for early detection of lumpy skin disease in cattle

Livestock farming serves a crucial role in global food security and sustainability, yet infections like Lumpy Skin Disease (LSD) present huge difficulties. LSD reduces the marketability and productivity of cattle, but it is not fatal. Current preventive measures that rely on routine veterinary testing can be time-consuming and miss early signs of infection, requiring a more effective management strategy first works This article presents an IoT-enabled intelligent system to prevent LSD in animals. The system uses a smart wearable sensor device to collect real-time health information including body temperature, heart rate and 3-axis motion data. Advanced analysis of data is considered to identify early diagnosis of potential LSD infection results based on deviations from normal parameters, this data prevents further infection within livestock by supporting immediate isolation. On a cloud-based online platform, and integrated into the mobile application provides a comprehensive experience and timely alerts to provide potential health issues. Furthermore, the IoT features of the system enable farmers and veterinarians to quickly guide and optimize disease control strategies, improving biosecurity protocols and overall livestock health management. This creative approach represents a significant improvement over traditional methods, providing a cost-effective, accurate and real-time solution for animal health monitoring and management.

Analysis of spatio-temporal variability of temperature, relative humidity and temperature humidity index in Naivasha Sub-county, Kenya

The aim of the study was to understand the environmental conditions impacting the reproductive performance of Sahiwal cattle, focusing on temperature, relative humidity, and relative humidity index (RHI). These factors significantly influence reproductive rates, highlighting the importance of spatio-temporal variability studies for informed herd management decisions. This study analyzed climate variability trends from 1998 to 2019 at the Kenya Agricultural and Livestock Research Organization (KALRO) Dairy Research Institute (DRI)-Naivasha, in Malewa Ward, Naivasha Sub-County, Kenya. Daily minimum and maximum temperature (T) and relative humidity (RH) data sourced from the Kenya Meteorological Department were utilized. The Temperature-Humidity Index (THI) was computed using the THI equation developed by Mader. Time series analyses, including the Mann-Kendall test and Sen's slope estimates, were employed to assess seasonal variability. Results indicated significant decreasing trends in temperature across all the four seasons studied (DJF, MAM, JJA, and SON), with rates of cooling at -0.0625, -0.0175, -0.0125, and -0.0342 °C per year, respectively. In contrast, relative humidity showed statistically significant increases in all seasons, with rates of +0.5977, +0.999, +1.4493, and +1.0499 for DJF, MAM, JJA, and SON, respectively. Seasonal THI exhibited significant decreases in DJF and JJA, potentially impacting livestock reproductive performance. These findings are crucial for policymakers to create adaptation and mitigation methods to address climate change effects on livestock and livelihoods.

Artificial intelligence for predicting arctic permafrost and active layer temperatures along the Alaskan North Slope

AbstractClimate change in Arctic regions poses a threat to permafrost stability, potentially leading to issues such as infrastructure damage, altered ecosystems, and increased greenhouse gas emissions. The challenge of monitoring permafrost temperatures and identifying critical environmental factors is accentuated by the scarcity of subsurface thermal data in remote Arctic locales. To alleviate this lack of soil thermal data we set out to combine in situ observed soil temperatures with MERRA-2 reanalysis data and Machine Learning (ML) to develop local and season-specific subsurface soil temperature predictions in Alaska. First, reanalysis features are selected based on surface energy budget physics. Coupled with Julian calendar dates, reanalysis features are preprocessed and then fed to the ML prediction model. We conducted a series of computational experiments and tested the results against performance benchmarks to determine the optimal prediction models, length of look-back period of model inputs, and the training set size. Six conventional ML models (e.g., GBDT, RF, SVR) and five statistical baselines (e.g. ARIMA) using in situ time series of soil temperature data ranging from 0 - 1 m depth are considered for the prediction task. The models are trained using in situ soil temperatures at depths between 0 and 1 m, spanning 16+ years, from field sites at Deadhorse and Toolik Lake, Alaska. All prediction performances are assessed using root mean squared error (RMSE) and mean absolute error (MAE). Results show that locally trained ML models can estimate shallow soil temperatures with an average error of $$RMSE=1.308^{\circ }$$ R M S E = 1 . 308 ∘ C.