#1Spatiotemporal vegetation variations and projections driven by atmosphere-ocean oscillations at multiple time scales: a case study in Gansu, China Qing He¹; Kwok Pan Chun¹; Xicai Pan 2; Liang Chen³; Pinyu Fan¹ 1 Department of Geography, Hong Kong Baptist University, Hong Kong, China 2 State Key Laboratory of Soil and Sustainable Agriculture, Institute of Soil Science, Chinese Academy of Sciences, Nanjing, China 3 Key Laboratory of Regional Climate Environment for Temperate East Asia, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China#2Content ☐ Research outline ☐ Research regions ☐ Vegetation variations ☐ Climate drivers ☐ Model assessment ☐ Vegetation projections#3Research outline Future vegetation projection Monsoons Webster and Yang Monsoon (WYM) Most important Regional climate change Central Pacific El Nino (CP) + CMIP5: climate models SST variability North Pacific El Nino (NP) Land use change Water CO2 fertilization Vegetation impacts indicator Regional environmental change monitor NDVI#4Research region inner land of northwest China The terrain is inclined from south-west to north-east, with an elevation ranging from 598 to 5602 m The climate in Gansu is complicated, involving cold desert climate (Bwk), cold semiarid climate (Bsk), temperate continental climate (Dwb) and cool continental climate (Bwc), based on Köppen Geiger climate classification system Owing to the arid climate, limited water resources and unsustainable exploitation, the desertification becomes more and more serious in the Gansu, which have greatly impeded the sustainable development of agriculture and economy in this region 32 34 36 38 Land Types 95 100 105 1 Crops, Mixed Farming 2 Short Grass 3 Evergreen Needleleaf Trees 4 Deciduous Needleleaf Trees 5 Deciduous Broadleaf Trees 6 Evergreen Broadleaf Trees 7 Tall Grass 8 Desert 9 Tundra 10 Irrigated Crops 11 Semidesert 12 Ice Caps and Glaciers 13 Bogs and Marshes 14 Inland Water 15 Ocean 16 Evergreen Shrubs 17 Deciduous Shrubs 18 Mixed Forest 19 Forest/Field Mosaic Figure 1. Land types with meteorological stations of Gansu, China.#5Vegetation variations 44°N The spatial distribution of the mean annual NDVI in Gansu exhibited high value in southeast but low value in northwest during the period of 2001 to 2017, suggesting high spatial variability in its distribution pattern. NDVI was found to be increased at a rate around 0.03 per year in southeast Gansu, but no significant changes in other regions Seasonal clusters of NDVI were found along both latitude and longitude. 42°N 40°N 38°N 36°N 34°N Figure 2. The mean NDVI (a) and NDVI trend (b) between 2001 and 2017. The NDVI along latitude(a) and longitude (d). For the trend map, only significant trends with 95% significance level are presented. (a) mean NDVI 44°N (b) NDVI trend 0.04 0.9 0.8 42°N 0.7 27 0.6 40°N 0.5 38°N 0.4 333 0.3 36°N 0.2 34°N 0.1 0.03 0.02 0.01 0 -0.01 -0.02 -0.03 32°N 0 32°N -0.04 92°E 96°E 100°E 104°E 108°E 92°E 96°E 100°E 104°E 108°E Longitude Latitude (c) NDVI along latitude 0.3 44 42 40 38 36 32 34 0.2 0.1 888 32 0 2002 2004 2006 2008 2010 2012 2014 2016 2018 (d) NDVI along longitude 0.3 110 105 100 95 0 2002 2004 2006 2008 2010 2012 2014 2016 2018 Time (years) 0.2 0.1#6Climate drivers The seasonal WYM and decadal NP had significant positive contributions to the vegetation coverage over the whole Gansu, especially the southeast region. Despite not statistically significant, the interannual CP negatively contributed to the vegetation variation over the most part of the Gansu, especially the northwest region, but positively to the southeast region. The vegetation coverage is mainly concentrated in the southeast Gansu, which is positively affected by three climate factors. NVDI WYM NVDI CP NVDI NP -0.3 -0.1 0.0 0.1 0.2 0.3 -0.03 -0.01 0.01 0.03 -0.015 -0.005 0.005 0.015 NVDI WYM NVDI CP NVDI NP Figure 3. The NDVI regressed by WYM, CP and NP. The left are regressed coefficient, while the right are only for significant coefficients. -0.3 -0.1 0.0 0.1 0.2 0.3 -0.03 -0.01 0.01 0.03 -0.015 -0.005 0.005 0.015#7Model assessment Model Table 1. CMIP5 climate models, institution and spatial resolution. Institution In this study, we used 5 climate models from the Coupled Model Intercomparison Project Phase 5 (CMIP5) with climate variables outputs available for two scenarios (RCP4.5 and RCP8.5), as summarized in Table 1. ACCESS1 Commonwealth Scientific and Industrial Research Organization (CSIRO) and Bureau of Spatial resolution (LonxLat) 1.875°x1.25° Meteorology (BOM), Australia GFDL Geophysical Fluid Dynamics Laboratory 2°x2.5° (GFDL), New Jersey bcc Beijing Climate Center (bcc), China ~2.8°x2.8° The Nash-Sutcliffe model efficiency (NSE) is used for model assessment. Meteorological Administration, China CNRM Centre National de Recherches ~1.4°x1.4° N Σ-1(xim-xi)² 2 M'et'eorologiques/Centre Europ´een de NSE = 1 2 Σ (Χ - Χο) the closer the NSE value to 1, the more reliable the estimation is. IPSL و MPI Recherche et For-mation Avanc´ees en Calcul Scientifique, France Institut Pierre-Simon Laplace (IPSL), France Max Planck Institute (MPI) for 3.75°x1.875° 1.875°x1.875° Meteorology, Germany#8Model assessment City Lon Lat (a) NSE under RCP45 Dunhuang (b) NSE under RCP85 94.71 40.13 Gaotai 99.84 Dunhuang 0.81 0.80 0.86 0.87 0.80 0.83 39.14 Dunhuang 0.81 0.80 0.83 0.83 0.80 0.84 0.9 Jiayuguan 98.28 39.78 0.9 Gaotai 0.86 0.84 0.87 0.90 0.85 0.87 Jiuquan Gaotai 0.86 0.84 0.84 0.88 98.5 0.84 0.89 39.71 0.8 Lanzhou 103.73 0.8 36.03 Jiayuguan 0.30 0.10 0.40 0.34 0.25 0.22- Tianshui Jiayuguan 0.16 0.41 0.33 0.19 0.23- 105.69 34.6 0.7 0.7 Wuwei 102.61 37.94 Jiuquan 0775 0.75 0.80 0.80 0.81 0.80 Jiuquan 0.75 0.77 0.77 0.80 0.81 Yumen 97.58 39.81 0.6 0.6 Zhangye 100.46 38.93 Lanzhou -0.64 0.58 0.70 0.62 0.65- Lanzhou 0.59 0.57 0.67 0.69 0.59 0.65- Xian 108.95 34.27 0.5 0.5 Tianshui 0.55 0.54 0.63 0.63 0.57 0.58- Wuwei 0.69 0274 0.77 Yumen 0.49 0.42 0.57 Under both RCP45 and RCP85, models have best performances in Dunhuang and Gaotai, and least in Jiayuguan and Zhangye 074 0.77 Xian. Generally, IPSL performs the best for all cities under both RCP45 and RCP85. Models have different performance for different cities. Xian 0.29 02 0.34 0.37 0.30 ACCESS1 CNRM Figure 5. The NSE between different modelled and satellite-based NDVI in different cities under RCP4.5 and RCP8.5. Tianshui 0.49 0.51 0.61 0.59 0.55 0.58 0.4 0.4 0.77 0.77 0.76- Wuwei 0.70 0.77 0.78 0.3 0.3 0.59 0.49 0.50 Yumen 0.47 0.44 0.2 0.81 0.82 0.81 0.81 bcc MPI IPSL GFDL Zhangye 173 0.76 0.77 0.1 0.35 Xian 0.31 0.17 0.33 0 bcc MPI IPSL GFDL CNRM ACCESS1 0.53 0.54 0.49 0.50 0.2 0.78 0.79 0.81 0.1 0.40 0.26 0.29- 0#9Vegetation projections 40°N Since most plants flourish in the summer and die in the winter, there is little change in vegetation throughout the year. In the vegetation growing season (i.e., summer July), two modelled NDVI showed a significant decline after 2020 for both RCP4.5 and RCP8.5 in Gansu. In winter (January), the NDVI is increasing under the RCP4.5, but decreases under the RCP8.5, suggesting the more greenhouse gas may harm the vegetation growth in Gansu. Latitude Latitude 35°N 40°N 35°N 200 km 100 mi 95°E 200 km IPSL under RCP45 in winter 100°E 105°E Longitude IPSL under RCP45 in summer 100 mi 95°E 100°E 105°E Longitude 4.73x104 -1.00×10-4 40°N ■decreasing ■increasing [4.97×10-4 Latitude 35°N -4.18×10-4 40°N decreasing increasing Latitude 35°N IPSL under RCP85 in winter 200 km 100 mi 95°E 100°E 105°E Longitude 事 IPSL under RCP85 in summer 200 km 100 mi 95°E 100°E 105°E Longitude -2.72x10-5 -1.43× 10-4 decreasing increasing 2.60x10-5 -1.94x104 decreasing ■increasing Figure 5. The bubble map of projected NDVI variations during 2020 and 2099 using RCP4.5 and RCP8.5 by IPSL model in different cities.