Drought impacts carbon and water fluxes of terrestrial ecosystems, which are strongly coupled. However, the magnitudes of response of carbon and water fluxes to drought are dependent on many processes, which are more complex than previously expected. Southern China experienced regional climatic perturbation events in the past decade and a two-year drought in 2009–2010. We used a terrestrial ecosystem model coupled with remotely sensed observations and metrological data to simulate the variations of net primary productivity (NPP), evapotranspiration (ET), and water-use efficiency (WUE) (i.e., NPP/ET) in south-western China during the period 2001–2010. Using the standard precipitation index (SPI) classifying different drought stresses, we also quantified the effect of drought on the ecosystem by comparing changes in modelled estimates of monthly WUE, NPP and ET under normal (i.e., baseline) and drought conditions (i.e., 2009 and 2010). The results indicated that NPP and ET showed synchronized declines in drought periods, with time-lag effects. Furthermore, drought-induced NPP decline was larger than ET reduction. An increasing trend in WUE from the moderate to extreme drought classes occurred not only in baseline conditions but also in drought conditions. Especially in the extreme drought period (January, 2010), WUE for the forest ecosystem typically showed a positive response to drought, indicating a drought-resilient forest ecosystem. Our study has important implications for understanding climate extreme effects on the carbon and water cycle of the forest ecosystem.

Developments in hyperspectral remote sensing techniques during the last decade have enabled the use of narrowband indices to evaluate the role of forest ecosystem variables in estimating carbon (C) fluxes. In this study, narrowband bio-indicators derived from EO-1 Hyperion data were investigated to determine whether they could capture the temporal variation and estimate the spatial variability of forest C fluxes derived from eddy covariance tower data. Nineteen indices were divided into four categories of optical indices: broadband, chlorophyll, red edge, and light use efficiency. Correlation tests were performed between the selected vegetation indices, gross primary production (GPP), and ecosystem respiration (Re). Among the 19 indices, five narrowband indices (Chlorophyll Index RedEdge 710, scaled photochemical reflectance index (SPRI)*enhanced vegetation index (EVI), SPRI*normalized difference vegetation index (NDVI), MCARI/OSAVI[705, 750] and the Vogelmann Index), and one broad band index (EVI) had R-squared values with a good fit for GPP and Re. The SPRI*NDVI has the highest significant coefficients of determination with GPP and Re (R² = 0.86 and 0.89, p < 0.0001, respectively). SPRI*NDVI was used in atmospheric inverse modeling at regional scales for the estimation of C fluxes. We compared the GPP spatial patterns inversed from our model with corresponding results from the Vegetation Photosynthesis Model (VPM), the Boreal Ecosystems Productivity Simulator model, and MODIS MOD17A2 products. The inversed GPP spatial patterns from our model of SPRI*NDVI had good agreement with the output from the VPM model. The normalized difference nitrogen index was well correlated with measured C net ecosystem exchange. Our findings indicated that narrowband bio-indicators based on EO-1 Hyperion images could be used to predict regional C flux variations for Northeastern China’s temperate broad-leaved Korean pine forest ecosystems.

Forest phenology is sensitive to climate change, and its responses affect many land surface processes, resulting in a feedback effect on climate change. Human activities have been the main driver of climate change’s long-term shifts in temperature and weather patterns. Forest phenology, understood as the timing of the annual cycles of plants, is extremely sensitive to changes in climate. Quantifying the responses of temperate forest phenology under an elevational range of topographic conditions that mimic climate change is essential for making effective adaptive forest ecosystem management decisions. Our study utilized the Google Earth Engine (GEE), gap filling, and the Savitzky–Golay (GF-SG) algorithm to develop a long-time series spatio-temporal remote sensing data fusion. The forest phenology characteristics on the north slope of Changbai Mountain were extracted and analyzed annually from 2013 to 2022. Our study found that the average start of the growing season (SOS) on the north slope of Changbai Mountain occurred between the 120th–150th day during the study period. The end of the growing season (EOS) occurred between the 270th–300th day, and the length of the growing season (LOS) ranged from the 110th–190th day. A transect from the northeast to southwest of the study area for a 10-year study period found that SOS was delayed by 39 d, the EOS advanced by 32 d, and the LOS was gradually shortened by 63 d. The forest phenology on the north slope of Changbai Mountain showed significant topographic differentiations. With an increase of 100 m in altitude, the mean SOS was delayed by 1.71 d (R² = 0.93, p < 0.01). There were no obvious trends in EOS variation within the study area altitude gradient. LOS decreased by 1.23 d for each 100 m increase in elevation (R² = 0.90, p < 0.01). Forests on steep slopes had an earlier SOS, a later EOS, and a longer LOS than forests on gentle slopes. For each degree increase in slope, SOS advanced by 0.12 d (R² = 0.53, p = 0.04), EOS was delayed by 0.18 d (R² = 0.82, p = 0.002), and the LOS increased by 0.28 d (R² = 0.78, p = 0.004). The slope aspect had effects on the EOS and the LOS but had no effect on the SOS. The forest EOS of the south aspect was 3.15 d later than that of the north aspect, and the LOS was 6.47 d longer. Over the 10-year study period, the phenology differences between the north and south aspects showed that the LOS difference decreased by 0.85 d, the SOS difference decreased by 0.34 d, and the EOS difference decreased by 0.53 d per year. Our study illustrates the significance of the coupling mechanism between mountain topography and forest phenology, which will assist our future understanding of the response of mountain forest phenology to climate change, and provide a scientific basis for further research on temperate forest phenology.

Background: One of the scientific challenges of understanding climate change has been determining the important drivers and metrics of global carbon (C) emissions and C cycling in tropical, subtropical, boreal, subarctic, and temperate peatlands. Peatlands account for 3% of global land cover, yet contain a major reservoir of 550 gigatons (Gt) of soil C, and serve as C sinks for 0.37 Gt of carbon dioxide (CO₂) a year. In the United States, temperate peatlands are estimated to store 455 petagrams of C (PgC). There has been increasing interest in the role of wildfires in C cycling and altering peatlands from C sinks to major C sources. We estimated above- and below-ground C emissions from the Pains Bay Fire, a long-duration wildfire (112 days; 18,329 ha) that burned a coastal peatland in eastern North Carolina, USA. Results: Soil C emissions were estimated from pre- and post-burn Light Detection and Ranging (LIDAR) soil elevation data, soils series and C content mapping, remotely sensed soil burn severity, and post-burn field surveys of soil elevation. Total above-ground C emissions from the fire were 2,89,579 t C and 214 t C ha⁻¹ for the 10 vegetation associations within the burn area perimeter. Above-ground sources of C emissions were comprised of litter (69,656 t C), shrub (1,68,983 t C), and foliage (50,940 t C). Total mean below-ground C emissions were 5,237,521 t C, and ranged from 2,630,529 to 8,287,900 t C, depending on organic matter content of different soil horizons within each of the 7 soil series. The mean below-ground C emissions within the burn area were 1,595.6 t C ha⁻¹ and ranged from 629.3 to 2511.3 t C ha⁻¹. Conclusions: In contrast to undisturbed temperate peatlands, human induced disturbances of the natural elevation gradient of the peatland has resulted in increased heterogeneity of floristic variation and assemblages that are a product of the spatial and temporal patterns of the water table level and the surface wetness across peatlands. Human induced changes in surface hydrology and land use influenced the fuel characteristics of natural vegetation and associated soils, thus influencing wildfire risk, behavior, and the resulting C emissions.

The photochemical reflectance index (PRI) has been suggested as an indicator of light use efficiency (LUE), and for use in the improvement of estimating gross primary production (GPP) in LUE models. Over the last two decades, solar-induced fluorescence (SIF) observations from remote sensing have been used to evaluate the distribution of GPP over a range of spatial and temporal scales. However, both PRI and SIF observations have been decoupled from photosynthesis under a variety of non-physiological factors, i.e., sun-view geometry and environmental variables. These observations are important for estimating GPP but rarely reported in the literature. In our study, multi-angle PRI and SIF observations were obtained during the 2018 growing season in a maize field. We evaluated a PRI-based LUE model for estimating GPP, and compared it with the direct estimation of GPP using concurrent SIF measurements. Our results showed that the observed PRI varied with view angles and that the averaged PRI from the multi-angle observations exhibited better performance than the single-angle observed PRI for estimating LUE. The PRI-based LUE model when compared to SIF, demonstrated a higher ability to capture the diurnal dynamics of GPP (the coefficient of determination (R²) = 0.71) than the seasonal changes (R² = 0.44), while the seasonal GPP variations were better estimated by SIF (R² = 0.50). Based on random forest analyses, relative humidity (RH) was the most important driver affecting diurnal GPP estimation using the PRI-based LUE model. The SIF-based linear model was most influenced by photosynthetically active radiation (PAR). The SIF-based linear model did not perform as well as the PRI-based LUE model under most environmental conditions, the exception being clear days (the ratio of direct and diffuse sky radiance > 2). Our study confirms the utility of multi-angle PRI observations in the estimation of GPP in LUE models and suggests that the effects of changing environmental conditions should be taken into account for accurately estimating GPP with PRI and SIF observations.