Search

• Discrepancies exist with different data in phenology extractions. • High temporal resolution data has higher accuracies in capturing SOS. • Double logistic method can minimize the extraction difference. The vegetation phenology is a commonly used indicator signaling vegetation responses to global changes. Monitoring vegetation phenology at a regional and global scale needs to rely to remote sensing data, for which multiple sources of datasets and extraction methods have been developed. To be efficient, remote sensing data with coarse temporal resolution is conventionally preferred in exploring vegetation phenology patterns at the continental or global scale. As fine temporal resolution data is increasingly available, effects of their temporal resolution on our analysis are still elusive. In this study, we applied several commonly utilized vegetation phenology extraction methods on two different temporal resolution MODIS NDVI data and compared their performances on the Tibetan Plateau (TP). The results showed there were certain discrepancies in the magnitude, trends and spatial patterns of extracted phenological parameters between the two datasets and among the different extraction methods. Generally, the phenological parameters derived from fine temporal resolution NDVI (MOD09A1) displayed later start of growing season (SOS), earlier end of growing season (EOS), shorter length of growing season (LOS), and were more accurate in capturing vegetation SOS compared with the coarse temporal resolution NDVI data (MOD13A2). The double logistic method can minimize the differences of extracted SOS or EOS between the two temporal resolution datasets. The findings of this study would improve accuracies of applying remote sensing data on monitoring vegetation dynamics and advance our understanding on vegetation responses to climate changes. [ABSTRACT FROM AUTHOR]

• Temporal change in relationship between NDVI and diurnal temperature was explored. • R NDVI-TMX and R NDVI-TMN showed distinct variations between mid- and high latitudes. • Weakening effect of TMX on vegetation growth was more apparent than that of TMN. • TRENDY models didn't capture dynamic relation of vegetation growth to TMX or TMN. Global warming has boosted vegetation growth to a large extent, but this stimulation effect has significantly weakened in recent years. Among the set of possible driving forces, the asymmetric daytime and nighttime warming effect has been largely neglected. To improve our understanding on the relationship between vegetation growth and global warming, this study tries to attribute the respective effects of daytime and nighttime temperature on vegetation growth and reveal their temporal trends in the extratropical Northern Hemisphere (30–90 °N). The results showed there had been significant warming trends in growing season maximum (TMX, 0.37 °C per decade) and minimum temperatures (TMN, 0.38 °C per decade) during 1982–2015, especially in high latitudes of the NH. Under the asymmetric diurnal warming, the effects of TMX and TMN on normalized difference vegetation index (NDVI) exhibited distinct temporal variations between mid- (<55 °N) and high latitudes (>55 °N). The positive correlation between NDVI and TMX (R NDVI-TMX) weakened in high latitudes, as well as the negative correlation between NDVI and TMN (R NDVI-TMN). However, the R NDVI-TMX and R NDVI-TMN changed little in mid-latitudes. Moreover, the weakening effect of TMX on NDVI was more apparent than that of TMN in high latitudes. The area with significantly (p < 0.1) positive R NDVI-TMX and significantly (p < 0.1) negative R NDVI-TMN both shrank from 1982–1998 to 1999–2015, with prior (15.97%) twice the latter one (7.09%) in shrunk area in high latitudes. With regard to vegetation type, decline in area with significantly (p < 0.1) positive R NDVI-TMX and negative R NDVI-TMN was extremely obvious in savannas, deciduous broadleaf forests and deciduous needleleaf forests. Besides, we also disclosed the poor performance of ecosystem process models in capturing the dynamic relationship between vegetation growth and diurnal temperature, which might be caused by their totally relying on diurnal mean temperature, instead of daytime and nighttime temperature, in forcing the models. This study can further advance our understandings on ecosystem responses to climate warming, and efficiently improve performance of ecosystem models. [ABSTRACT FROM AUTHOR]

• The first study to use high-resolution images to survey pika hole size and area. • The first study to explore pika hole distribution pattern at a landscape scale. • Effects of pika hole on plant growth in the surrounding varies with pika hole size. • Vegetation coverage affects pika hole density in alpine meadow. The pika (Ochotona curzoniae) hole is an important landscape feature in the Tibetan Plateau (TP) grasslands, and it indicates grassland degradation levels due to the destruction caused by pika burrowing activities on grasslands. However, no studies have ever explored landscape patterns of pika holes and their effects on adjacent vegetation coverage. Taking meadow grasslands in Northern Tibet as an example, this study gathered unmanned aerial vehicle (UAV) images and explored landscape patterns of pika holes and their effects on grass coverage in the surroundings. The performances of two classification methods, including the decision tree classification based on Fully Constrained Least Squares (FDC) and the object-oriented classification (OBC) were compared in recognizing sizes and shapes of pika holes. The results showed that: (1) The object-oriented classification exhibits higher classification accuracy in identifying pika holes. (2) The average size of pika holes in the study area is 0.01 m2 and they exhibit clustered distribution patterns. The average distance between any two nearest pika hole patches is 0.79 m. (3) It presents a significant quadratic relationship between the number of pika holes and grass coverage. (4) The average effective distance of pika holes on the surrounding grass coverage is 20 cm. The findings of this study can provide guidelines for pika control and improve grassland management on the TP. [ABSTRACT FROM AUTHOR]

• Phenology length dominantly influences carbon sequestration in Xizang Plateau. • The dominant role of phenology length enhances along an elevational gradient. • The plateau may play an increasingly crucial role in mitigating future climate change. Interannual variations in terrestrial carbon sequestration are jointly controlled by vegetation phenology and physiological activity. However, it remains unclear how this joint control varies with elevation in alpine ecosystems. To explore the relationships among these factors in Xizang Plateau grassland during 1981–2014, we used the remotely sensed total enhanced vegetation index (EVI) during the growing season (EVI GST), growing season length (GSL), and maximum EVI during the growing season (EVI max) as indicators of annual carbon sequestration, phenology, and photosynthetic activity, respectively. Using these indicators, we investigated the effects of phenology and photosynthetic activity on annual carbon sequestration along an elevation gradient. We found that EVI GST increased, indicating an improvement in carbon sequestration, under prolonged GSL and enhanced EVI max. The trend of a prolonged GSL increased with elevation by 0.039 day year−1 per 1000 m (p < 0.001). The interannual variation in EVI GST was better explained by the interannual variation in GSL (standardized regression coefficient, SRC: 0.88; linear R2: 0.89; partial R2: 0.85) than by that in EVI max (SRC: 0.12; linear R2: 0.33; partial R2: 0.09). This stronger contribution of GSL than of EVI max became more pronounced with increasing elevation. The increase in GSL was more effective than the resistance of environmental constraints of the maximum photosynthetic activity on strengthening carbon sequestration under more extreme environments at higher elevations. These findings increase our understanding of phenological effects on carbon sequestration in alpine ecosystems. Given the projected climate warming and wetting in the future, our study suggests that carbon sequestration will increase due to climate-induced prolongation of GSL across Xizang Plateau grassland. The plateau may play an increasingly crucial role in mitigating future climate change. [ABSTRACT FROM AUTHOR]

Although forest landscape models (FLMs) have benefited greatly from ongoing advances of computer technology and software engineering, computing capacity remains a bottleneck in the design and development of FLMs. Computer memory overhead and run time efficiency are primary limiting factors when applying forest landscape models to simulate large landscapes with fine spatial resolutions and great vegetation detail. We introduce LANDIS PRO 6.0, a landscape model that simulates forest succession and disturbances on a wide range of spatial and temporal scales. LANDIS PRO 6.0 improves on existing forest landscape models with two new data structures and algorithms (hash table and run-length compression). The innovative computer design enables LANDIS PRO 6.0 to simulate very large (>108 ha) landscapes with a 30-m spatial resolution, which to our knowledge no other raster forest landscape models can do. We demonstrate model behavior and performance through application to five nested forest landscapes with varying sizes (from 1 million to 100 million 0.09-ha cells) in the southern Missouri Ozarks. The simulation results showed significant and variable effects of changing spatial extent on simulated forest succession patterns. Results highlighted the utility of a model like LANDIS PRO 6.0 that is capable of efficiently simulating large landscapes and scaling up forest landscape processes to a common regional scale of analysis. The programming methodology presented here may significantly advance the development of next generation of forest landscape models. [Copyright &y& Elsevier]