Bias Temperature Research Articles

Large-Scale Surface Air Temperature Bias in Summer over the CONUS and its Relationship to Tropical Central Pacific Convection in the UFS Prototype 8

Large-Scale Surface Air Temperature Bias in Summer over the CONUS and its Relationship to Tropical Central Pacific Convection in the UFS Prototype 8

Threshold voltage instability of SiC MOSFETs under very‐high temperature and wide gate bias

AbstractThreshold voltage (VTH) instability affects the reliability of silicon carbide (SiC) MOSFETs. In this article, the influence of gate bias (VGS) and high temperature on VTH instability is investigated under wide VGS and very‐high temperature range (150°C to 275°C). The degradation mechanism of gate oxide under coupling of different electrical and thermal stress is revealed. When the device is subjected to +VGS, the relationship between VTH shift and temperature is not monotonic and there is a temperature turning point. This is mainly related to the capture/release and generation of electron traps. However, the VTH shift increases sharply and gate oxide breakdown for the device bias at +35 V, which is related to Fowler–Nordheim (F–N) tunnelling effect. For a large −VGS, the VTH shift is very small. Moreover, when the negative VGS decreases, the VTH shift is positively correlated with temperature, which may result from the activation and charge exchange of hole traps. However, the influence of moving ions changes the temperature dependence of VTH shift for the device bias at −30 V. Finally, the devices of different manufacturers are studied and similar change patterns are found. This finding provides guidance for the further application of SiC MOSFETs under high temperatures and different voltages.

Greenland summer blocking characteristics: an evaluation of a high-resolution multi-model ensemble

AbstractAtmospheric blocking is a phenomenon that can lead to extreme weather events over a large region, yet its causes are not fully understood. Global climate models show limitations in representing Northern Hemisphere blocking, especially its frequency, and decadal variability in Greenland blocking in summer in the recent decades. In this study we evaluate the ability of high-resolution (HighResMIP) Earth System Models (ESMs) to simulate summer blocking over the Greenland area, using different but complementary methods to describe the characteristics of blocking. We find that the HighResMIP ensemble can reproduce the spatial pattern of Greenland blocking events, albeit with systematic biases, and capture the relative frequencies of the main blocking patterns: namely the wave breaking structure, North Atlantic ridge, and omega-type blocking. However, the HighResMIP ensemble fails to simulate the observed temporal variations of Greenland blocking index (GB2) and the extremely high values of daily GB2 observed in recent decades. In addition, we do not find clearly superior representation of blocking features from higher-resolution in HighResMIP models compared with lower-resolution models. We also find large sea surface temperature (SST) biases over the North Atlantic and seas surrounding Greenland, and biases in moisture transport over the North Atlantic toward Greenland, especially over the western flank of blocking areas, which might together contribute to model biases in the representation of blocking magnitude.

Medium‐range predictability of temperature extremes and biases in Rossby‐wave amplitude

AbstractThis study investigates the medium‐range predictability of warm and cold extremes in the Northern Hemisphere and the role that upper‐tropospheric circulation biases play in this regard. Deterministic ERA5 reforecasts for the period 1979–2019 are evaluated based on the ERA5 reanalysis of the respective period, thus providing a large sample for verification and bias identification. The predictability of temperature extremes at 850 hPa is assessed based on the Gilbert Skill Score and other metrics and is shown to exhibit regional and seasonal variations. Summer is generally characterized by lower forecast skill scores than winter for both warm and cold extremes. Moreover, cold extremes in summer have slightly lower skill scores than warm extremes, while the opposite is true in winter. Biases in the frequency of temperature extremes are, to some extent, consistent with biases in mean temperature and indicate an underestimation in the total amount of extremes for much of the hemisphere in summer. Associated with the latter, biases also emerge in the standard deviation of the daily temperature distribution, with the summer values being largely underestimated over most of the hemisphere. The role of upper‐tropospheric circulation in these biases is then assessed by verifying the representation of Rossby‐wave packet (RWP) properties. It is found that the amplitude of RWPs is systematically underestimated in most of the hemisphere in summer, while it is overestimated in many parts of the midlatitudes in winter. Overall, the results suggest that the underestimation of RWP amplitude in summer hinders the medium‐range predictability of temperature extremes in the explored retrospective and operational forecasts. Although operational European Centre for Medium‐Range Weather Forecasts (ECMWF) forecasts gradually improve between 2013 and 2022 in terms of the 850‐hPa temperature and 300‐hPa RWP amplitude absolute errors, the aforementioned summer biases remain qualitatively similar.

Surface Warming Constraint Projects Less Permafrost Thawing in High Mountain Asia

AbstractReliable projections of permafrost change are crucial for estimating permafrost carbon loss. However, potential model biases in surface air temperature may yield unrealistic projections of future permafrost area. Here, by leveraging the emergent relationship between equilibrium climate sensitivity and projected changes in mean annual air temperature over High Mountain Asia (HMA), we mitigate the overestimated local warming rates and excessive thawing of permafrost associated with the “hot model” problem in models participating in the Coupled Model Intercomparison Project phase 6. After constraint, permafrost area over HMA will reduce by 37%, 64% and 99% in 2081–2100 relative to present‐day under the SSP1‐2.6, SSP2‐4.5 and SSP5‐8.5 scenarios, respectively. In contrast, the unconstrained projections tend to overestimate the loss of permafrost area by nearly 10% under the low and mid‐emission scenarios due to the overestimation of local warming rates. These findings are crucial for policymaking and provide valuable insights into global permafrost projection.

Antarctic sea ice surface temperature bias in atmospheric reanalyses induced by the combined effects of sea ice and clouds

Sea-ice surface temperature from atmospheric reanalysis has been used as an indicator of ice melt and climate change. However, its performance in atmospheric reanalyses is not fully understood in Antarctica. Here, we quantified biases in six widely-used reanalyses using satellite observations, and found strong and persistent warm biases in most reanalyses examined. Further analysis of the biases revealed two main culprits: incorrect cloud properties, and inappropriate sea-ice representation in the reanalysis products. We found that overestimated cloud simulation can contribute more than 4 K warm bias, with ERA5 exhibiting the largest warm bias. Even in reanalysis with smaller biases, this accuracy is achieved through a compensatory relationship between relatively lower cloud fraction bias and overestimated sea ice insulation effect. A dynamic downscaling simulation shows that differences in sea-ice representation can contribute a 2.3 K warm bias. The representation of ice concentration is the primary driver of the spatial distribution of biases by modulating the coupling between sea ice and clouds, as well as surface heat conduction. The lack of a snow layer in all reanalyses examined also has an impact on biases.

Qingfei Zhisou oral liquid alleviates fever-induced inflammation by regulating arachidonic acid and lysophospholipids metabolism and inhibiting hypothalamus transient receptor potential ion channels expression.

To explore how Qingfei Zhisou oral liquid (, QFZS) adjusts body temperature bias and the interaction of inflammatory factors levels and metabolomic differences. Dry yeast was subcutaneously injected at 10 mL/kg to establish the pyrexia model. We randomly divided 60 Sprague-Dawley rats into five groups: control, model, positive, low dose of QFZS and high dose of QFZS. Inflammatory proteins were evaluated by Western blotting and immunohistochemistry. For the examination of the endogenous metabolites, enzyme linked immunosorbent assay and ultra-high-performance liquid chromatography high-resolution mass spectrometry were employed. QFZS significantly reduced rats' body temperature within 6 h after dry yeast injection and reduced the secretion of the arginine vasopressin, cyclic adenosine monophosphate, prostaglandin E-2, tumor necrosis factor-α (TNF-α), interleukin-6 (IL-6), and interleukin-1β in serum. Meanwhile, we identified 41 metabolites between the model and QFZS groups, including arachidonic acid and lysophospholipids. QFZS restored normal arachidonic acid levels. Based on the differential metabolite enrichment analysis, QFZS's anti-inflammatory and anti-pyrexia effects might be related to the inflammatory pathway regulated by transient receptor potential. Additionally, QFZS treatment reduced transient receptor potential melastatin 2 ion channel expression and affected TNF-α, heat shock protein 70, and cyclooxygenase-2 expression in the hypothalamus. QFZS exerts its regulatory effects on fever by regulating the metabolism of lysophospholipids and arachidonic acid and the regulation of inflammation via transient receptor potential ion channels channels.

Substantial Cold Bias During Wintertime Cold Extremes in the Southern Cascadia Region in Historical CMIP6 Simulations

AbstractGlobal climate models often simulate atmospheric conditions incorrectly due to their coarse grid resolution, flaws in their dynamics, and biases resulting from parameterization schemes. Here we document a bias in the magnitude and extent of minimum temperature extremes in the CMIP6 model ensemble, relative to ERA5. The bias is present in the southern Cascadia region (i.e., Pacific Northwestern United States and southwestern British Columbia, Canada, spanning from the coast to the Rocky Mountains), with some models showing a bias magnitude in excess of −10°C in the first percentile of daily winter minimum temperature. The sea level pressure pattern for these events is similar in CMIP6 models and ERA5, showing high anomalies in the Northeast Pacific that are indicative of an atmospheric blocking pattern and consequently more northerly flow. Though this atmospheric blocking pattern is typically concurrent with cold winter temperatures across much of North America, Rocky and Cascade mountain ranges prevent the cold air from reaching the southern Cascadia region as confirmed by the observations and reanalysis. Our results suggest that the bias in CMIP6 minimum temperatures is a result of unresolved topography in the Rockies and Cascade mountain ranges, such that the terrain does not adequately block cold air advection from reaching the southern Cascadia region.

The influence of El Niño on springtime synoptic-scale precipitation extremes in Southeastern China: insights from CMIP6 model simulations

This study focuses on El Niño impacts on springtime extreme precipitation in Southeastern China (SEC) by comparing observations with data from the Coupled Model Intercomparison Project phase 6 (CMIP6) historical runs. Observational and simulated results suggest that synoptic-scale temperature advection patterns over East Asia (EA) are closely associated with extreme precipitation in SEC, encompassing the Pearl River Basin (PRB), Yangtze River Basin (YRB), and Huaihe River Basin (HRB). Based on this, we introduce a temperature advection index (TAI) tailored to capture the cold-warm temperature advection dipole, which shows a significant positive correlation with SEC precipitation. Both observations and CMIP6 indicate that TAI-related circulations, characterized by upper-level synoptic-scale waves and a north–south oriented temperature gradient over EA, are conducive to extreme precipitation in northern PRB (NPRB)–YRB–HRB. However, the TAI-related synoptic-scale activities have a lesser impact on extreme precipitation in southern PRB (SPRB), as these disturbances mainly affect the mid-latitude weather. Further investigation reveals that during boreal spring following El Niño, 85% of extreme events in YRB–HRB are associated with positive TAI values, compared to 76% under climatological conditions. However, such a change in the association with TAI is not evident in CMIP6 simulations. From observations, atmospheric baroclinicity along the East Asian westerly jet is enhanced during El Niño, which promotes the development of TAI-related synoptic-scale disturbances. In contrast, CMIP6 models struggle to reproduce these observed baroclinicity signals during El Niño. This challenge arises from the background westerly jet bias and mean-state cold tongue bias in tropical Pacific temperature in models.

Ultrahigh Thermal Rectification at Small Temperature Difference Achieved by Near‐Field Thermal Photon Manipulation

AbstractThermal diodes, which allow heat to flow in a preferential direction, are an essential component of thermal circuitry and will find applications in many fields ranging from thermal photon controls to quantum information. The quality of a thermal diode is defined by the thermal rectification ratio (TRR). Here, an ultrahigh TRR exceeding 105 at a small temperature difference (1–5 K) is presented based on near‐field thermal photon (NFTP) manipulation, by utilizing asymmetric gap‐variated dual terminals based on graphene/Si heterostructures and materials with contrasting thermal expansion coefficients. Analyses of the photon tunneling probability show that the colossal TRR comes from coupled and decoupled graphene surface plasmon polaritons under opposite temperature biases, remaining robust across a wide operating temperature range. Further enhancement can be achieved by tuning the Fermi level of graphene. This work demonstrates the significance of asymmetry in NFTP manipulation for thermal diode performance, achieving an ultrahigh TRR, which is at least three orders higher than the state‐of‐the‐art structures at the same small temperature difference (1–5 K), and is comparable to state‐of‐the‐art electronic diodes.

Assessment of the Southern Ocean Sea Surface Temperature Biases in CMIP5 and CMIP6 Models

Assessment of the Southern Ocean Sea Surface Temperature Biases in CMIP5 and CMIP6 Models

The effect of temperature variation on the transient response of RF PIN diode limiters for very high frequency applications

AbstractThis work presents the effect of temperature change on the capacitance of silicon PIN diodes and the resulting change in performance of RF limiters at very high frequency (VHF). Device temperatures were varied between −25 ºC and 100 ºC, with small‐signal parameters (including device capacitance) extracted at regular temperature increments and bias voltages from −20 Vdc to +3 Vdc using a multi‐bias parameter extraction method. It was found that the junction capacitance of the four PIN diodes under investigation increases with temperature, as expected from carrier lifetime behaviour, while results also confirmed prior observations of an inverse relationship between forward‐biased series resistance and temperature. Devices were subsequently tested in two different limiter topologies through high‐power transient measurements. It was found that the combination of increased capacitance and decreased resistance with increasing temperature increases the transient spike leakage and decreases the flat leakage of a limiter. It was also concluded that, for VHF, an anti‐parallel topology provides the best performance over a wide range of temperatures.

Integrating physics-based WRF atmospheric variables and machine learning algorithms to predict snowfall accumulation in Northeast United States

Accurate snowfall prediction is crucial for enhancing preparedness and resilience in the Northeast United States during winter weather events. This study introduces a novel approach that integrates Machine Learning (ML) models with atmospheric variables from the Weather Research and Forecasting (WRF) model to improve snowfall forecasts in the region. The significance lies in bridging the gap between physics-based Numerical Weather Prediction (NWP) models and the versatility of ML models, offering a promising advancement in winter storm predictions. Atmospheric variables related to 32 winter storms simulated by WRF, were used to feed Random Forest (RF) and Extreme Gradient Boosting (XGBoost) algorithms to predict 24 h accumulations of snowfall using the National Snowfall Analysis (NSA) product as reference. The comprehensive results revealed that the integrated approach provided more accurate snowfall prediction than the WRF and Air Force Weather Agency (AFWA) diagnostic, but faced challenges in the precision estimated by under-dispersion of the results. The WRF/XGBoost reduced the RMSE by 10.34 %, whereas WRF/RF reduced RMSE by 9.72 %, compared to WRF/AFWA. Similarly, WRF/XGBoost and WRF/RF increased the correlation coefficient by 12 % and 11 %. The most important variables in both ML algorithms were liquid water equivalent precipitation (LWE), snow ratio, wet-bulb temperature, temperature, and humidity at different heights, pressure tendency and wind speed, validating their ability to discern crucial factors influencing snowfall. Specific cases highlight the integrated approach’s effectiveness in addressing challenges faced by traditional diagnostics, including LWE overestimation and warm temperature bias. However, limitations emerge in capturing storms characterized by anomalous atmospheric conditions and high snowfall values, most likely attributed to underrepresentation in the training data. By highlighting strengths and acknowledging limitations, this integrated approach holds the potential for improved snowfall prediction for future storms in the region compared to the traditional approach.

A novel temperature drift compensation method based on LSTM for NMR sensor

Nuclear Magnetic Resonance (NMR) sensors have become one of the best choices for angular velocity sensors in future Inertial Navigation Systems (INS) due to their small size and high precision. The performance of NMR sensors can be affected by temperature changes, resulting in temperature drift. Therefore, temperature compensation is essential. Accurate temperature compensation is challenging due to the lack of direct temperature observations in the core sensitive area (cell) and the complex temperature drift model. Considering the time correlation of temperatures inside and outside the cell, a new temperature drift compensation method based on the Long-Short Term Memory (LSTM) network is proposed. The memory characteristics of the LSTM enable it to fully utilize current and previous data to learn complex models. The main contributions of this study are as follows: (1) the mechanism of temperature drift in NMR sensors was analyzed; (2) a multi-time scale input–output model for temperature changes, temperature change rates, and ambient temperature versus sensor output bias was established; (3) a temperature drift model identification method based on LSTM was designed. Experimental results show that, compared to traditional polynomial fitting and memory-free methods, the proposed method improves temperature compensation accuracy by 26.0% to 47.0%. This method effectively reduces the adverse impact of temperature changes on the NMR sensor.

Enhanced exchange bias in all-oxide heterostructures with cation-ordered ferrimagnetic double-perovskite

The realization and control of exchange bias (EB) are highly desirable for spintronic applications. All-oxide heterostructures comprised of ferromagnetic and antiferromagnetic/multiferroic oxides provide an ideal platform to enable the electric-field control of EB, promising for energy-efficient memory and logic devices. However, the low block temperature (TB) and small bias field (HEB) hinder further advances towards room-temperature applications. Here, we report an alternative approach to enhance the interface-induced EB by using ferrimagnetic double-perovskite with B-site cation ordering. In heterostructures comprised of double-perovskite Sr2FeReO6 (SFRO) and LaFeO3 (LFO), a high TB (about 250 K) and large HEB are observed, which is significantly larger than the counterparts with LFO and ferromagnetic oxides. Further analysis suggests that the cation-ordering and ferrimagnetic spin structure of the double-perovskite could contribute significantly to the enhanced exchanged bias when interfacing with G-type antiferromagnets. Our results open a new avenue for developing all-oxides heterostructures for future magnetic technologies.

Assessing the Tropospheric Temperature and Humidity Simulations in CMIP3/5/6 Models Using the AIRS Obs4MIPs V2.1 Data

AbstractIn this study, the Atmospheric Infrared Sounder (AIRS) Observations for Model Intercomparison Projects (Obs4MIPs) V2.1 tropospheric air temperature, specific humidity, and relative humidity data are utilized to evaluate the global tropospheric temperature and humidity simulations in the fully coupled global climate models from the Coupled Model Intercomparison Project phases 3, 5, and 6 (CMIP3, CMIP5, and CMIP6), and possible simulation improvement in CMIP6 models in comparison to CMIP3 and CMIP5 models. Our analyses indicate that all three phases of CMIP models share similar tropospheric air temperature, specific humidity, and relative humidity biases in their multi‐model ensemble means relative to AIRS. Cold biases up to 4 K and positive relative humidity biases up to 20% are found in the free troposphere almost globally with maxima over the mid‐latitude storm tracks. Warm biases up to 2 K are seen over the Southern Ocean in the lower troposphere. Positive specific and relative humidity biases exist over the off‐equatorial oceans while negative specific and relative humidity biases are seen near the equator in the tropical free troposphere, which are related to the double‐intertropical convergence zone bias in the models. Both the air temperature and specific humidity biases are important to the relative humidity biases except in the tropical free troposphere where the specific humidity biases dominate. The tropospheric air temperature, specific humidity, and relative humidity biases are reduced from CMIP3 to CMIP5 and to CMIP6 at almost all pressure levels except at 300 hPa for specific humidity and in the boundary layer for relative humidity.

Elevation-dependent biases of raw and bias-adjusted EURO-CORDEX regional climate models in the European Alps

Data from the EURO-CORDEX ensemble of regional climate model simulations and the CORDEX-Adjust dataset were evaluated over the European Alps using multiple gridded observational datasets. Biases, which are here defined as the difference between models and observations, were assessed as a function of the elevation for different climate indices that span average and extreme conditions. Moreover, we assessed the impact of different observational datasets on the evaluation, including E-OBS, APGD, and high-resolution national datasets. Furthermore, we assessed the bi-variate dependency of temperature and precipitation biases, their temporal evolution, and the impact of different bias adjustment methods and bias adjustment reference datasets. Biases in seasonal temperature, seasonal precipitation, and wet-day frequency were found to increase with elevation. Differences in temporal trends between RCMs and observations caused a temporal dependency of biases, which could be removed by detrending both observations and RCMs. The choice of the reference observation datasets used for bias adjustment turned out to be more relevant than the choice of the bias adjustment method itself. Consequently, climate change assessments in mountain regions need to pay particular attention to the choice of observational dataset and, furthermore, to the elevation dependence of biases and the increasing observational uncertainty with elevation in order to provide robust information on future climate.

Spatial and Size Distributions of Ti(C5H7O2)2[(CH3)2CHO]2 Mist Particles in a Tubular Furnace for Conformal and Uniform Deposition of Amorphous TiO2 Thin Films

The spatial and size distributions of titanium diisopropoxide bisacetylacetonate [(C5H7O2)2[(CH3)2CHO]2, also known as Ti(acac)2(OiPr)2] mist, diluted in CH3OH, are investigated in a tubular furnace using atmospheric‐pressure mist chemical vapor deposition (mist CVD). The focus is on the deposition of amorphous (a)‐TiO2 films with tubular furnace temperature and mesh bias as variables. When the furnace temperature reaches 350 °C, the number density of mist particles increases without significant changes in their size distribution, leading to a higher film deposition rate. Further, the deposition rate and average size of the mist particles with lower adhesion coefficient decrease with increasing spatial distance from the furnace inlet. Furthermore, applying a mesh bias results in an increase in the maximum number density of mist particles with a narrower size distribution; however, the overall film deposition rate decreases. These variations are attributed to the chemical reactivity of the mist precursors produced by pyrolysis and mesh bias. The fine mist precursors, which are strongly charged, coordinate with CH3OH and CHO groups through solvation, enhancing their chemical stability and lifetime. This process yields a dense and rigid a‐TiO2 network, improving the junction properties at the a‐TiO2/c‐Si interface.

Evaluation of the coupling of EMACv2.55 to the land surface and vegetation model JSBACHv4

Abstract. We present the coupling of the Jena Scheme for Biosphere–Atmosphere Coupling in Hamburg version 4 (JSBACHv4) to the ECHAM/MESSy Atmospheric Chemistry (EMAC) model. With JSBACH, the soil water bucket model in EMAC is replaced by a diffusive hydrological transport model for soil water that includes water storage and infiltration in five soil layers, preventing soil from drying too rapidly and reducing biases in soil temperature and moisture. A three-layer soil scheme is implemented, and phase changes in water in the soil are considered. The leaf area index (LAI) climatology in EMAC has been substituted with a phenology module calculating the LAI. Multiple land cover types are included to provide a state-dependent surface albedo, which accounts for the absorption of solar radiation by vegetation. Plant net primary productivity, leaf area index and surface roughness are calculated according to the plant functional types. This paper provides a detailed evaluation of the new coupled model based on observations and reanalysis data, including ERA5/ERA5-Land datasets, Global Precipitation Climatology Project (GPCP) data and Moderate Resolution Imaging Spectroradiometer (MODIS) satellite data. Land surface temperature (LST), terrestrial water storage (TWS), surface albedo (α), net top-of-atmosphere radiation flux (RadTOA), precipitation (precip), leaf area index (LAI), fraction of absorbed photosynthetic active radiation (FAPAR) and gross primary productivity (GPP) are evaluated in particular. The strongest correlation (r) between reanalysis data and the newly coupled model is found for LST (r=0.985, with an average global bias of −1.546 K), α (r=0.947, with an average global bias of −0.015) and RadTOA (r=0.907, with an average global bias of 3.56 W m−2). Precipitation exhibits a correlation with the GPCP dataset of 0.523 and an average global bias of 0.042 mm d−1. The LAI optimisation yields a correlation of 0.637 with observations and a global mean deviation of −0.212. FAPAR and GPP exemplify two of the many additional variables made available through JSBACH in EMAC. FAPAR and observations show a correlation of 0.663, with an average global difference of −0.223, while the correlation for GPP and observations is 0.564 and the average global difference is −0.001 kg carbon km−1. Benefiting from the numerous added features within the simulated land system, the representation of soil moisture is improved, which is critical for vegetation modelling. This improvement can be attributed to a general increase in soil moisture and water storage in deeper soil layers and a closer alignment of simulated TWS with observations, mitigating the previously widespread problem of soil drought. We show that the numerous newly added components strongly improve the land surface, e.g. soil moisture, TWS and LAI, while surface parameters, such as LST, surface albedo or RadTOA, which were mostly prescribed according to climatologies, remain similar. The coupling of JSBACH brings EMAC a step closer towards a holistic comprehensive Earth system model and extends its versatility.

Projection of snowfall and precipitation phase changes over the Northwest China based on CMIP6 multimodels

The change of precipitation phase could have profound influence on water cycle and ecological environment in Northwest China (NWC). It has been observed that snowfall is inclined to transition into rainfall under climate warming. However, the evolution process of precipitation phase in the future over the NWC remains poorly understanding. In this study, the performance of snowfall simulated by Coupled Model Intercomparison Project Phase 6 (CMIP6) is evaluated by results of observation-based rain/snow separation, subsequently analyzing the changes in snowfall and precipitation phase under different climate scenarios over the NWC. The results indicated that most of CMIP6 models overestimate snowfall while underestimate the Snowfall to Precipitation Ratio (SPR) over the NWC. The performance of CMIP6 models is poorer in high-altitude mountainous region but relatively more robust in low-altitude areas. Superior performance of five models are selected for analysing changes of snowfall and precipitation phase in the future. Under the SSP1-2.6 scenario, the trend of changes in snowfall is not obviously in the future. Conversely, snowfall is projected to decline at rates of −1.04 mm·10a-1 and −2.94 mm·10a-1 under the SSP2-4.5 and SSP5-8.5 scenarios, respectively, with significant declines in high-altitude zones. The SPR will decrease at rates of −0.19 %·10a-1, −0.64 %·10a-1, and −1.50 %·10a-1 under SSP1-2.6, SSP2-4.5, and SSP5-8.5 scenario, respectively. The magnitudes in changes of snowfall and SPR show significant differences across sub-regions in the future, with low-altitude areas exhibiting upward trends in snowfall and SPR under certain conditions. Winter snowfall in the future shows an increased trend, while declines are observed in spring and autumn snowfall overall, with notable disparities between scenarios. The correlation between snowfall biases and biases of temperature and precipitation indicate that temperature is a pivotal factor contributing to precipitation phase bias. The findings are helpful for enhancing the simulation precision of hydrological processes and providing scientific strategies for future water resource management over the NWC.