基于随机森林模型的黄土高原草地净初级生产力时空格局及未来演变趋势模拟

doi:10.13866/j.azr.2023.01.13

干旱区研究 ›› 2023, Vol. 40 ›› Issue (1): 123-131.doi: 10.13866/j.azr.2023.01.13

基于随机森林模型的黄土高原草地净初级生产力时空格局及未来演变趋势模拟

刘欢欢¹(),陈印¹,刘悦¹,刚成诚^2,³()

1.西北农林科技大学草业与草原学院，陕西杨凌 712100
2.西北农林科技大学水土保持研究所，陕西杨凌 712100
3.中国科学院水利部水土保持研究所，陕西杨凌 712100

收稿日期:2022-05-26 修回日期:2022-08-13 出版日期:2023-01-15 发布日期:2023-02-24
通讯作者: 刚成诚. E-mail: gangcc@ms.iswc.ac.cn
作者简介:刘欢欢（1999-），男，硕士研究生，研究方向为草地生态遥感. E-mail: liuhh75@nwafu.edu.cn
基金资助:
陕西省自然科学基金(2021JQ-171);青海省防灾减灾重点实验室开放基金(QFZ-2021-Z06);国家自然科学基金项目(31602004);国家科技基础条件平台建设项目(2005DKA32300)

Simulation of spatial pattern and future trends of grassland net primary productivity in the Loess Plateau based on random forest model

LIU Huanhuan¹(),CHEN Yin¹,LIU Yue¹,GANG Chengcheng^2,³()

1. College of Grassland Agriculture, Northwest A & F University, Yangling 712100, Shaanxi, China
2. Institute of Soil and Water Conservation, Northwest A & F University, Yangling 712100, Shaanxi, China
3. Institute of Soil and Water Conservation, Chinese Academy of Sciences and Ministry of Water Resources, Yangling 712100, Shaanxi, China

Received:2022-05-26 Revised:2022-08-13 Online:2023-01-15 Published:2023-02-24

摘要/Abstract

摘要：

准确估算草地净初级生产力（Net Primary Productivity，NPP）是了解草地生态系统碳循环过程和科学评估草地对气候变化响应与适应机制的关键。以黄土高原草地生态系统为研究对象，基于1788个草地生物量数据及19个环境因子数据（包含气候、植被、土壤和地形因子），利用随机森林模型模拟了2002—2020年黄土高原草地NPP的时空动态，并估算了共享社会经济路径（Shared Socioeconomic Pathway，SSPs）4个未来气候情景下，黄土高原草地NPP的未来演变趋势。结果表明：（1）随机森林模型的模拟精度较好，可用于黄土高原草地NPP估算；（2）黄土高原草地NPP在空间上整体呈现“东南高西北低”的空间分布特征，年均值为276.55 g C·m^-2，其中陕西关中地区草地NPP最高；（3） 2002—2020年，黄土高原草地NPP总体呈现增加趋势，其中55.01%的区域草地NPP增加，主要集中在陕西关中地区、甘肃西部地区以及山西北部地区；（4）在气候暖湿化背景下，到本世纪末，黄土高原草地NPP均呈增加趋势，其中SSP585情景下草地NPP增加最多，SSP126情景下增加最少。利用随机森林模型能够较好模拟黄土高原草地NPP时空格局及未来演变趋势，为黄土高原草地生态系统保护及可持续发展提供数据支持。

关键词: 草地生态系统, 净初级生产力, 随机森林, 共享社会经济路径, 黄土高原

Abstract:

The accurate estimation of grassland net primary productivity (NPP) is crucial to understanding the carbon cycle of grassland ecosystems and their adaption to climate change. Based on 1788 grassland biomass data and 19 environmental factors (climate, vegetation, soil, and topographic factors), we simulated the spatiotemporal dynamics of grassland NPP in the Loess Plateau from 2002 to 2020 using the random forest (RF) model. The future trends of grassland NPP under four future climate scenarios of shared socioeconomic pathways were estimated. Results showed that (1) the RF model had a good accuracy, which indicated that RF can be used to estimate grassland NPP in the Loess Plateau; (2) grassland NPP in the Loess Plateau exhibited a “high in southeastern and low in northwestern” pattern, with a mean value of 276.55 g C·m^-2·a^-1. The highest grassland NPP was observed in Guanzhong Plain of Shaanxi; (3) the grassland NPP in the Loess Plateau showed an overall increasing trend during 2002-2020. Regions experiencing an increase in grassland NPP accounted for 55.01% of the total land area, which is mainly located in Guanzhong Plain, western Gansu, and northern Shanxi; (4) under the wetter and warmer climate, grassland NPP in the Loess Plateau will continually increase by the end of this century. Grassland NPP will increase the most under the SSP585 scenario and the least under the SSP126 scenario. RF can be used to simulate the temporal and spatial trends of grassland NPP in the Loess Plateau. The results provide data support for the protection and sustainable development of grassland ecosystem in the Loess Plateau.

Key words: grassland ecosystems, net primary productivity (NPP), random forest, Shared Socioeconomic Pathway (SSPs), Loess Plateau

刘欢欢, 陈印, 刘悦, 刚成诚. 基于随机森林模型的黄土高原草地净初级生产力时空格局及未来演变趋势模拟[J]. 干旱区研究, 2023, 40(1): 123-131.

LIU Huanhuan, CHEN Yin, LIU Yue, GANG Chengcheng. Simulation of spatial pattern and future trends of grassland net primary productivity in the Loess Plateau based on random forest model[J]. Arid Zone Research, 2023, 40(1): 123-131.

图/表 8

图1

表1

图2

图3

图4

图5

图6

图7

参考文献 44

[1]	刚成诚, 王钊齐, 杨悦, 等. 近百年全球草地生态系统净初级生产力时空动态对气候变化的响应[J]. 草业学报, 2016, 25(11): 1-14.
	[Gang Chengcheng, Wang Zhaoqi, Yang Yue, et al. The NPP spatiotemporal variation of global grassland ecosystems in response to climate change over the past 100 years[J]. Acta Prataculturae Sinica, 2016, 25(11): 1-14.]
[2]	Kang L, Han X, Zhang Z, et al. Grassland ecosystems in China: Review of current knowledge and research advancement[J]. Philosophical Transactions of the Royal Society B: Biological Sciences, 2007, 362(1482): 997-1008. doi: 10.1098/rstb.2007.2029
[3]	Sun J, Du W. Effects of precipitation and temperature on net primary productivity and precipitation use efficiency across China’s grasslands[J]. Giscience & Remote Sensing, 2017, 54(6): 881-897.
[4]	王翀. 三江源区高寒草地净初级生产力模拟研究[D]. 兰州: 兰州大学, 2013.
	[Wang Chong. Study on Simulation Methods of Alpine Grassland Net Primary Productivity in Three Rivers Source Region of Tibetan Plateau, China[D]. Lanzhou: Lanzhou University, 2013.]
[5]	马文红, 方精云, 杨元合, 等. 中国北方草地生物量动态及其与气候因子的关系[J]. 中国科学: 生命科学, 2010, 40(7): 632-641.
	[Ma Wenhong, Fang Jingyun, Yang Yuanhe, et al. Dynamics of grassland biomass in northern China and its relationship with climatic factors[J]. Scientia Sinica(Vitae), 2010, 40(7): 632-641.]
[6]	王莺, 夏文韬, 梁天刚, 等. 基于MODIS植被指数的甘南草地净初级生产力时空变化研究[J]. 草业学报, 2010, 19(1): 201-210.
	[Wang Yuan, Xia Wentao, Liang Tiangang, et al. Spatial and temporal dynamic changes of net primary product based on MODIS vegetation index in Gannan grassland[J]. Acta Prataculturae Sinica, 2010, 19(1): 201-210.]
[7]	Guo D, Song X, Hu R, et al. Grassland type-dependent spatiotemporal characteristics of productivity in Inner Mongolia and its response to climate factors[J]. Science of the Total Environment, 2021, 775: 145617. doi: 10.1016/j.scitotenv.2021.145617
[8]	张赟鑫, 郝海超, 范连连, 等. 中亚草地NPP时空动态及其驱动因素研究[J]. 干旱区研究, 2022, 39(3): 698-707.
	[Zhang Yunxin, Hao Haichao, Fan Lianlian, et al. Study on spatio-temporal dynamics and driving factors of NPP in Central Asian grassland[J]. Arid Zone Research, 2022, 39(3): 698-707.]
[9]	Zhang C, Zhang Y, Li J. Grassland productivity response to climate change in the Hulunbuir Steppes of China[J]. Sustainability, 2019, 11(23): 6760-6775. doi: 10.3390/su11236760
[10]	Jung M, Reichstein M, Margolis H A, et al. Global patterns of land-atmosphere fluxes of carbon dioxide, latent heat, and sensible heat derived from eddy covariance, satellite, and meteorological observations[J]. Journal of Geophysical Research-Biogeosciences, 2011, 116(G3): 1566-1582.
[11]	Yao Y, Wang X, Li Y, et al. Spatiotemporal pattern of gross primary productivity and its covariation with climate in China over the last thirty years[J]. Global Change Biology, 2018, 24(1): 184-196. doi: 10.1111/gcb.13830 pmid: 28727222
[12]	Sun Y, Feng Y, Wang Y, et al. Field-based estimation of net primary productivity and its above and belowground partitioning in global grassland[J]. Journal of Geophysical Research: Biogeosciences, 2021, 126(11): 6472-6496.
[13]	Li C, Dou T, Wang Y, et al. A method for quantifying the impacts of human activities on net primary production of grasslands in Northwest China[J]. Remote Sensing, 2021, 13(13): 2479-2497. doi: 10.3390/rs13132479
[14]	李传华, 孙皓, 王玉涛, 等. 基于机器学习估算青藏高原多年冻土区草地净初级生产力[J]. 生态学杂志, 2020, 39(5): 1734-1744.
	[Li Chuanhua, Sun Hao, Wang Yutao, et al. Estimation of grassland net primary productivity in permafrost of Qinghai-Tibet Plateau based on machine learning[J]. Chinese Journal of Ecology, 2020, 39(5): 1734-1744.]
[15]	Fu B. Soil erosion and its control in the loess plateau of China[J]. Soil Use and Management, 1989, 5(2): 76-82. doi: 10.1111/j.1475-2743.1989.tb00765.x
[16]	韩庆功, 彭守璋. 黄土高原潜在自然植被空间格局及其生境适宜性[J]. 水土保持学报, 2021, 35(5): 188-193, 203.
	[Han Qinggong, Peng Shouzhang. Spatial pattern and habitat suitability of potential natural vegetation in the Loess Plateau[J]. Journal of Soil and Water Conservation, 2021, 35(5): 188-193, 203.]
[17]	Liu F, Yan H, Gu F, et al. Net primary productivity increased on the loess plateau following implementation of the grain to green program[J]. Journal of Resources and Ecology, 2017, 8(4): 413-421. doi: 10.5814/j.issn.1674-764x.2017.04.014
[18]	史晓亮, 杨志勇, 王馨爽, 等. 黄土高原植被净初级生产力的时空变化及其与气候因子的关系[J]. 中国农业气象, 2016, 37(4): 445-453.
	[Shi Xiaoliang, Yang Zhiyong, Wang Xinshuang, et al. Spatial and temporal variation of net primary productivity and its relationship with climate factors in the Chinese Loess Plateau[J]. Chinese Journal of Agrometeorology, 2016, 37(4): 445-453.]
[19]	刘洋洋, 王倩, 杨悦, 等. 黄土高原草地净初级生产力时空动态及其影响因素[J]. 应用生态学报, 2019, 30(7): 2309-2319. doi: 10.13287/j.1001-9332.201907.002
	[Liu Yangyang, Wang Qian, Yang Yue, et al. Spatial-temporal dynamics of grassland NPP and its driving factors in the Loess Plateau, China[J]. Chinese Journal of Applied Ecology, 2019, 30(7): 2309-2319.] doi: 10.13287/j.1001-9332.201907.002
[20]	Zhao G, Mu X, Wen Z, et al. Soil Erosion, Conservation, and eco-environment changes in the Loess Plateau of China[J]. Land Degradation & Development, 2013, 24(5): 499-510. doi: 10.1002/ldr.2246
[21]	Wang Y, Shao M, Zhu Y, et al. Impacts of land use and plant characteristics on dried soil layers in different climatic regions on the Loess Plateau of China[J]. Agricultural and Forest Meteorology, 2011, 151(4): 437-448. doi: 10.1016/j.agrformet.2010.11.016
[22]	孙锐, 陈少辉, 苏红波. 黄土高原不同生态类型NDVI时空变化及其对气候变化响应[J]. 地理研究, 2020, 39(5): 1200-1214. doi: 10.11821/dlyj020190399
	[Sun Rui, Chen Shaohui, Su Hongbo. Spatiotemporal variation of NDVI different ecotypes on the Loess Plateau and its response to climate change[J]. Geographical Research, 2020, 39(5): 1200-1214.] doi: 10.11821/dlyj020190399
[23]	Jia W, Liu M, Yang Y, et al. Estimation and uncertainty analyses of grassland biomass in Northern China: Comparison of multiple remote sensing data sources and modeling approaches[J]. Ecological Indicators, 2016, 60: 1031-1040. doi: 10.1016/j.ecolind.2015.09.001
[24]	Peng S, Ding Y, Liu W, et al. 1 km monthly temperature and precipitation dataset for China from 1901 to 2017[J]. Earth System Science Data, 2019, 11(4): 1931-1946. doi: 10.5194/essd-11-1931-2019
[25]	Li X, Xiao J. A Global, 0.05-degree product of solar-induced chlorophyll fluorescence derived from OCO-2, MODIS, and reanalysis data[J]. Remote Sensing, 2019, 11(5): 517-526. doi: 10.3390/rs11050517
[26]	Su Y, Li J, Zhi H, et al. Projected drought conditions in Northwest China with CMIP6 models under combined SSPs and RCPs for 2015-2099[J]. Advances in Climate Change Research, 2020, 11(3): 210-217. doi: 10.1016/j.accre.2020.09.003
[27]	Breiman L. Random forests[J]. Machine Learning, 2001, 45(1): 5-32. doi: 10.1023/A:1010933404324
[28]	刘畅, 张红, 张霄羽, 等. 半干旱地区矿区土地利用时空演变与预测[J]. 干旱区研究, 2022, 39(1): 292-300.
	[Liu Chang, Zhang Hong, Zhang Xiaoyu, et al. Spatio-temporal evolution and prediction of land use in semi-arid mining areas[J]. Arid Zone Research, 2022, 39(1): 292-300.]
[29]	Biau G, Scornet E. A random forest guided tour[J]. Test, 2016, 25(2): 197-227. doi: 10.1007/s11749-016-0481-7
[30]	Belgiu M, Dragut L. Random forest in remote sensing: A review of applications and future directions[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2016, 114: 24-31. doi: 10.1016/j.isprsjprs.2016.01.011
[31]	Ohlson J, Kim S. Linear valuation without OLS: The Theil-Sen estimation approach[J]. Review of Accounting Studies, 2015, 20(1): 395-435. doi: 10.1007/s11142-014-9300-0
[32]	Ullah S, You Q, Ali A, et al. Observed changes in maximum and minimum temperatures over China-Pakistan economic corridor during 1980-2016[J]. Atmospheric Research, 2019, 216: 37-51. doi: 10.1016/j.atmosres.2018.09.020
[33]	Zeng N, Ren X, He H, et al. Estimating grassland aboveground biomass on the Tibetan Plateau using a random forest algorithm[J]. Ecological Indicators, 2019, 102: 479-487. doi: 10.1016/j.ecolind.2019.02.023
[34]	魏建洲. 黄土高原草地植被变化及其驱动力分析[D]. 兰州: 兰州大学, 2020.
	[Wei Jianzhou. Grassland Vegetation Change and its Driving Force in the Loess Plateau of China[D]. Lanzhou: Lanzhou University, 2020.]
[35]	Li C, Wang Y, Wu X, et al. Reducing human activity promotes environmental restoration in arid and semi-arid regions: A case study in Northwest China[J]. Science of the Total Environment, 2021, 768: 144525. doi: 10.1016/j.scitotenv.2020.144525
[36]	Li C, Dou T, Wang Y, et al. A method for quantifying the impacts of human activities on net primary production of grasslands in Northwest China[J]. Remote Sensing, 2021, 13(13): 2479-2497. doi: 10.3390/rs13132479
[37]	Liu G, Shao Q, Fan J, et al. Change trend and restoration potential of vegetation net primary productivity in China over the past 20 years[J]. Remote Sensing, 2022, 14(7): 1634-1660. doi: 10.3390/rs14071634
[38]	刘铮, 杨金贵, 马理辉, 等. 黄土高原草地净初级生产力时空趋势及其驱动因素[J]. 应用生态学报, 2021, 32(1): 113-122. doi: 10.13287/j.1001-9332.202101.017
	[Liu Zheng, Yang Jingui, Ma Lihui, et al. Spatial-temporal trend of grassland net primary production and their driving factors in the Loess Plateau, China[J]. Chinese Journal of Applied Ecology, 2021, 32(1): 113-122.] doi: 10.13287/j.1001-9332.202101.017
[39]	杨丹, 王晓峰. 黄土高原气候和人类活动对植被NPP变化的影响[J]. 干旱区研究, 2022, 39(2): 584-593.
	[Yang Dan, Wang Xiaofeng. Contribution of climatic change and human activities to changes in net primary productivity in the Loess Plateau[J]. Arid Zone Research, 2022, 39(2): 584-593.]
[40]	Gang C, Zhao W, Zhao T, et al. The impacts of land conversion and management measures on the grassland net primary productivity over the Loess Plateau, Northern China[J]. Science of the Total Environment, 2018, 645: 827-836. doi: 10.1016/j.scitotenv.2018.07.161
[41]	Wu D, Zhao X, Liang S, et al. Time-lag effects of global vegetation responses to climate change[J]. Global Change Biology, 2015, 21(9): 3520-3531. doi: 10.1111/gcb.12945 pmid: 25858027
[42]	Gang C, Gao X, Peng S, et al. Satellite observations of the recovery of forests and grasslands in Western China[J]. Journal of Geophysical Research-Biogeosciences, 2019, 124(7): 1905-1922. doi: 10.1029/2019JG005198
[43]	Yu H, Wu Y, Niu L, et al. A method to avoid spatial overfitting in estimation of grassland above-ground biomass on the Tibetan Plateau[J]. Ecological Indicators, 2021, 125: 107450. doi: 10.1016/j.ecolind.2021.107450
[44]	Karlson M, Ostwald M, Reese H, et al. Mapping tree canopy cover and aboveground biomass in sudano-sahelian woodlands using Landsat 8 and random forest[J]. Remote Sensing, 2015, 7(8): 10017-10041. doi: 10.3390/rs70810017

类别	环境因子	解释	分辨率	数据来源
S	CLAY_S	0~30 cm黏粒含量	250 m	https://data.isric.org/
S	CLAY_T	30~100 cm黏粒含量	250 m
S	SILT_S	0~30 cm粉粒含量	250 m
S	SILT_T	30~100 cm粉粒含量	250 m
S	SAND_S	0~30 cm 砂粒含量	250 m
S	SAND_T	30~100 cm砂粒含量	250 m
S	SOC_S	0~30 cm土壤有机碳含量	250 m
S	SOC_T	30~100 cm土壤有机碳含量	250 m
T	DEM	数字高程模型	30 m	https://srtm.csi.cgiar.org/
T	Slope	坡度	30 m
A	TEM	年均温（2002—2020）	1000 m	http://loess.geodata.cn
A	TEM_4-10	4—10月均温（2002—2020）	1000 m
A	TMN	年最低温度（2002—2020）	1000 m
A	PRE	年降水量（2002—2020）	1000 m
A	PRE_4-10	4—10月降水量（2002—2020）	1000 m
A	ET	蒸散量（2002—2020）	500 m	https://modis.gsfc.nasa.gov
B	SIF	日光诱导叶绿素荧光（2002—2020）	0.05°	https://globalecology.unh.edu
B	FPAR	植被有效光合辐射吸收比例（2002—2020）	500 m	https://modis.gsfc.nasa.gov
B	NDVI	归一化植被指数（2002—2020）	1000 m	https://modis.gsfc.nasa.gov

基于随机森林模型的黄土高原草地净初级生产力时空格局及未来演变趋势模拟

Simulation of spatial pattern and future trends of grassland net primary productivity in the Loess Plateau based on random forest model

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 8

参考文献 44

相关文章 15

Metrics

本文评价

推荐阅读 0

[1]	赵雨琪, 魏天兴. 1990—2020年黄土高原典型县域植被覆盖变化及影响因素[J]. 干旱区研究, 2024, 41(1): 147-156.
[2]	刘一丹, 姚晓军, 李宗省, 胡家瑜. 气候变化和土地利用覆盖变化对河西地区植被净初级生产力的影响[J]. 干旱区研究, 2024, 41(1): 169-180.
[3]	李小雨, 贾科利, 魏慧敏, 陈睿华, 王怡婧. 基于随机森林算法的土壤含盐量预测[J]. 干旱区研究, 2023, 40(8): 1258-1267.
[4]	任丽雯, 王兴涛, 刘明春, 王大为. 石羊河流域植被净初级生产力时空变化及驱动因素[J]. 干旱区研究, 2023, 40(5): 818-828.
[5]	吕锦心, 梁康, 刘昌明, 张仪辉, 刘璐. 无定河流域土地覆被空间分异机制及相关水碳变量变化[J]. 干旱区研究, 2023, 40(4): 563-572.
[6]	薛智暄, 张丽, 王新军, 李永康, 张冠宏, 李沛尧. 古尔班通古特沙漠SMAP土壤水分产品降尺度分析[J]. 干旱区研究, 2023, 40(4): 583-593.
[7]	王怡恩, 饶良懿. 气候因素和人类活动对砒砂岩区植被净初级生产力的影响[J]. 干旱区研究, 2023, 40(12): 1982-1995.
[8]	吴雪晴, 张乐乐, 高黎明, 李炎坤, 刘轩辰. 青海湖流域NPP动态变化及驱动力[J]. 干旱区研究, 2023, 40(11): 1824-1832.
[9]	裴宏泽, 赵亚超, 张廷龙. 2000—2020年黄土高原NEP时空格局与驱动力[J]. 干旱区研究, 2023, 40(11): 1833-1844.
[10]	肖森天, 依力亚斯江·努尔麦麦提, 努尔比耶·穆合塔尔, 赵静, 阿迪莱·阿卜来提. 基于光学和雷达多源遥感的于田绿洲土壤盐渍化时空分析[J]. 干旱区研究, 2023, 40(1): 59-68.
[11]	方贺,严佩文,石见,康娟,刘海蓉,陈丹,罗继,徐栋. 阿克苏地区植被生态质量时空变化及其驱动机制[J]. 干旱区研究, 2022, 39(6): 1907-1916.
[12]	安彬,肖薇薇,朱妮,刘宇峰. 近60 a黄土高原地区降水集中度与集中期时空变化特征[J]. 干旱区研究, 2022, 39(5): 1333-1344.
[13]	龙志,孙颖琦,郎丽霞,陈兴鹏,张子龙,庞家幸. 黄土高原典型县域碳排放特征与时空格局——以庆城县为例[J]. 干旱区研究, 2022, 39(5): 1631-1641.
[14]	祝存兄,史飞飞,乔斌,张娟,陈国茜. 基于高分1号卫星数据的青海湖扩张及湖滨沙地变化特征分析[J]. 干旱区研究, 2022, 39(4): 1076-1089.
[15]	冯强,赵文武,段宝玲. 生态系统服务权衡强度与供需匹配度的关联性分析——以山西省为例[J]. 干旱区研究, 2022, 39(4): 1222-1233.