Temperature sensitivity of respiration of forest soils is important for its responses to climate warming and for the accurate assessment of soil carbon budget. The sensitivity of temperature (T_i) to soil respiration rate (R_s), and Q₁₀ defined by e^{10(lnRs−lna)/Ti} has been used extensively for indicating the sensitivity of soil respiration. The soil respiration under a larch (Larix gmelinii) forest in the northern Daxing’an Mountains, Northeast China was observed in situ from April to September, 2019 using the dynamic chamber method. Air temperatures (T_air), soil surface temperatures (T_0cm), soil temperatures at depths of 5 and 10 cm (T_5cm and T_10cm, respectively), and soil-surface water vapor concentrations were monitored at the same time. The results show a significant monthly variability in soil respiration rate in the growing season (April–September). The Q₁₀ at the surface and at depths of 5 and 10 cm was estimated at 5.6, 6.3, and 7.2, respectively. The Q_10@10 cm over the period of surface soil thawing (Q_{10@10 cm, thaw} = 36.89) were significantly higher than that of the growing season (Q_{10@10 cm, growth} = 3.82). Furthermore, the R_s in the early stage of near-surface soil thawing and in the middle of the growing season is more sensitive to changes in soil temperatures. Soil temperature is thus the dominant factor for season variations in soil respiration, but rainfall is the main controller for short-term fluctuations in respiration. Thus, the higher sensitivity of soil respiration to temperature (Q₁₀) is found in the middle part of the growing season. The monthly and seasonal Q₁₀ values better reflect the responsiveness of soil respiration to changes in hydrometeorology and ground freeze-thaw processes. This study may help assess the stability of the soil carbon pool and strength of carbon fluxes in the larch forested permafrost regions in the northern Daxing’an Mountains.

The responsiveness of four types of stem diameter shrinkage indicators to sap flow changes was compared under four air temperature and cloudiness conditions: above 0 °C below 80% cloud cover days; above 0 °C large percentage cloud cover days; low temperature below 80% cloud cover days; and low temperature large percentage cloud cover days. In this study, we investigated the effects of indicative functions of relatively easy-to-access stem diameter shrinkage on variation characteristics of sap flow. High-resolution-based stem diameter shrinkage is related to changes in tree moisture content. Stem diameter shrinkage indicators are adopted to confirm sap flow changes resulting from transpiration pull, which may enhance the power of stem diameter shrinkage as an index for tree bole moisture loss. After measuring stem diameter variations, the following stem diameter shrinkage indicators were calculated: maximum daily shrinkage, daily stem diameter increment, daily stem diameter variation, and tree water deficit-induced stem shrinkage (TWD). Sap flow was measured synchronously, and stem diameter shrinkage indicators were analyzed to confirm their responses to sap flow. TWD was positively correlated (r ≥ 0.317) with daily variations in sap flow and reached extremely significant levels (p ≤ 0.001) under all conditions. TWD and maximum daily shrinkage were able to better reflect the correlation between changes in stem diameter and sap flow on a daily scale, except large percentage cloud cover days with low temperatures. Changes in stem diameter had no correlation with sap flow during low temperature and large percentage cloud cover days. Among all stem diameter shrinkage indicators, TWD showed the highest correlation (r ≥ 0.601 and p ≤ 0.001) with sap flow under all conditions, except during large percentage cloud cover days with low temperatures. The stem diameter shrinkage indicators did not reflect sap flow changes during large percentage cloud cover days with low temperatures. The indicator that best reflected moisture loss of trees was TWD.

The response mechanism of the tree stem radius variation to hydro-thermal factors is complex and diverse. The changes of TWD (tree water deficit-induced stem shrinkage) and GRO (growth-induced irreversible stem expansion) are respectively driven by different factors, so that their responses to hydro-thermal factors are different. The stem radius variation and its matching hydro-thermal factors experimental data was measured and determined at 0.5 h time scale in larch (Larix gmelini Rupr.) forest of the Daxing’anling region of the most northeastern part of China. Response characteristics of the stem radius variation to hydro-thermal factors have been found by analyzing the data under different time windows. The stem radius variation mainly responded to the changes in precipitation and relative humidity. The main driving factors for TWD were sap flow density and solar radiation. The response of GRO to hydro-thermal factors was complex, varied a lot under different time scales. During the analysis of the response of tree radial growth, changes of the stem radius can be divided to TWD and GRO to implement separate studies on their responses to hydro-thermal factors. In this way, it becomes easier to discover the response of TWD under drought stress and the responding mechanism of GRO to hydro-thermal factors.

Mapping the distribution of coniferous forests is of great importance to the sustainable management of forests and government decision-making. The development of remote sensing, cloud computing and deep learning has provided the support of data, computing power and algorithms for obtaining large-scale forest parameters. However, few studies have used deep learning algorithms combined with Google Earth Engine (GEE) to extract coniferous forests in large areas and the performance remains unknown. In this study, we thus propose a cloud-enabled deep-learning approach using long-time series Landsat remote sensing images to map the distribution and obtain information on the dynamics of coniferous forests over 35 years (1985–2020) in the northwest of Liaoning, China, through the combination of GEE and U²-Net. Firstly, to assess the reliability of the proposed method, the U²-Net model was compared with three Unet variants (i.e., Resnet50-Unet, Mobile-Unet and U-Net) in coniferous forest extraction. Secondly, we evaluated U²-Net’s temporal transferability of remote sensing images from Landsat-5 TM, Landsat-7 ETM+ and Landsat-8 OLI. Finally, we compared the results obtained by the proposed approach with three publicly available datasets, namely GlobeLand30-2010, GLC_FCS30-2010 and FROM_GLC30-2010. The results show that (1) the cloud-enabled deep-learning approach proposed in this paper that combines GEE and U²-Net achieves a high performance in coniferous forest extraction with an F1 score, overall accuracy (OA), precision, recall and kappa of 95.4%, 94.2%, 96.6%, 95.5% and 94.0%, respectively, outperforming the other three Unet variants; (2) the proposed model trained by the sample blocks collected from a specific time can be applied to predict the coniferous forests in different years with satisfactory precision; (3) Compared with three global land-cover products, the distribution of coniferous forests extracted by U²-Net was most similar to that of actual coniferous forests; (4) The area of coniferous forests in Northwestern Liaoning showed an upward trend in the past 35 years. The area of coniferous forests has grown from 945.64 km² in 1985 to 6084.55 km² in 2020 with a growth rate of 543.43%. This study indicates that the proposed approach combining GEE and U²-Net can extract coniferous forests quickly and accurately, which helps obtain dynamic information and assists scientists in developing sustainable strategies for forest management.