Spatial non-stationary characteristics between grass yield and its influencing factors in the Ningxia temperate grasslands based on a mixed geographically weighted regression model

doi:10.1007/s11442-022-1986-5

Abstract

Abstract:

Spatial models are effective in obtaining local details on grassland biomass, and their accuracy has important practical significance for the stable management of grasses and livestock. To this end, the present study utilized measured quadrat data of grass yield across different regions in the main growing season of temperate grasslands in Ningxia of China (August 2020), combined with hydrometeorology, elevation, net primary productivity (NPP), and other auxiliary data over the same period. Accordingly, non-stationary characteristics of the spatial scale, and the effects of influencing factors on grass yield were analyzed using a mixed geographically weighted regression (MGWR) model. The results showed that the model was suitable for correlation analysis. The spatial scale of ratio resident-area index (PRI) was the largest, followed by the digital elevation model, NPP, distance from gully, distance from river, average July rainfall, and daily temperature range; whereas the spatial scales of night light, distance from roads, and relative humidity (RH) were the most limited. All influencing factors maintained positive and negative effects on grass yield, save for the strictly negative effect of RH. The regression results revealed a multiscale differential spatial response regularity of different influencing factors on grass yield. Regression parameters revealed that the results of Ordinary least squares (OLS) (Adjusted R²= 0.642) and geographically weighted regression (GWR) (Adjusted R²= 0.797) models were worse than those of MGWR (Adjusted R²= 0.889) models. Based on the results of the RMSE and radius index, the simulation effect also was MGWR > GWR > OLS models. Ultimately, the MGWR model held the strongest prediction performance (R² = 0.8306). Spatially, the grass yield was high in the south and west, and low in the north and east of the study area. The results of this study provide a new technical support for rapid and accurate estimation of grassland yield to dynamically adjust grazing decision in the semi-arid loess hilly region.

Key words: grass yield, spatial non-stationary, mixed geographically weighted regression model, temperate grassland, Ningxia

SONG Xiaolong, MI Nan, MI Wenbao, LI Longtang. Spatial non-stationary characteristics between grass yield and its influencing factors in the Ningxia temperate grasslands based on a mixed geographically weighted regression model[J].Journal of Geographical Sciences, 2022, 32(6): 1076-1102.

Figures/Tables 15

Figure 1

Figure 2

Table 1

Figure 3

Table 2

Table 3

Figure 4

Table 4

Table 5

Table 6

Figure 5

Table 7

Figure 6

Figure 7

Table 8

References 74

[1]	Akaike H T, 1974. A new look at the statistical model identification. IEEE Transactions on Automatic Control, 19(6): 716-723. doi: 10.1109/TAC.1974.1100705
[2]	Anderson G L, Hanson J D, Haas R H, 1993. Evaluating Landsat thematic mapper derived vegetation indices for estimating above-ground biomass on semiarid rangelands. Remote Sensing of Environment, 45(2): 165-175. doi: 10.1016/0034-4257(93)90040-5
[3]	Arshad A, Zhang W, Zhang Z, et al., 2021. Reconstructing high-resolution gridded precipitation data using an improved downscaling approach over the high altitude mountain regions of Upper Indus Basin (UIB). Science of The Total Environment, 784(2021): 147140. doi: 10.1016/j.scitotenv.2021.147140
[4]	Bai X, Wen Z, An S, et al., 2015. Evaluating sustainability of cropland use in Yuanzhou County of the Loess Plateau, China using an emergy-based ecological footprint. Plos One, 10(3): 1-10.
[5]	Bergen S, Sheppard L, Sampson P D et al., 2013. A national prediction model for PM_2.5 component exposures and measurement error-corrected health effect inference. Environmental Health Perspectives, 121(9): 1017-1025. doi: 10.1289/ehp.1206010 pmid: 23757600
[6]	Brunsdon C, Fotheringham A S, Charlton M E, 1996. Geographically weighted regression: A method for exploring spatial nonstationarity. Geographical Analysis, 28: 281-298. doi: 10.1111/j.1538-4632.1996.tb00936.x
[7]	Brunsdon C, Fotheringham A S, Charlton M E, 2002. Geographically weighted summary statistics: A framework for localised exploratory data analysis. Computers Environment & Urban Systems, 26(6): 501-524.
[8]	Cao X, Sun L, Zhao Z, et al., 2018. Application of MODIS remote sensing products in the estimation of grass yield in Sanjiang Source Area. Remote Sensing for Land & Resources, 30(4): 115-124. (in Chinese)
[9]	Cao Y N, Wu J S, Zhang X Z, et al., 2019. Dynamic forage-livestock balance analysis in alpine grasslands on the northern Tibetan Plateau. Journal of Environmental Management, 238(15): 352-359. doi: 10.1016/j.jenvman.2019.03.010
[10]	Chao L, Zhang K, Li Z, et al., 2018. Geographically weighted regression based methods for merging satellite and gauge precipitation. Journal of Hydrology, 558: 275-289. doi: 10.1016/j.jhydrol.2018.01.042
[11]	Chen Y, Zhu M K, Zhou Q, et al., 2021. Research on spatiotemporal differentiation and influence mechanism of urban resilience in China based on MGWR Model. International Journal of Environmental Research and Public Health, 18(3): 1056. doi: 10.3390/ijerph18031056
[12]	Cleveland W S, 1979. Robust locally weighted regression and smoothing scatterplots. Journal of the American Statistical Association, 74(368): 829-836. doi: 10.1080/01621459.1979.10481038
[13]	Dai Y L, Tang W, Wang S, et al., 2012. Monitoring report on natural grassland vegetation change after returning grazing land to grassland in southern mountainous area of Ningxia. Ningxia Journal of Agriculture and Forestry Science and Technology, 3(11): 118-119, 122. (in Chinese)
[14]	Dong R L, Jia W, Yu G H, 2018. Research progress on prediction models of grass yield and livestock carrying capacity of grassland. Acta Agrestia Sinica, 26(5): 1043-1051. (in Chinese)
[15]	Fotheringham A S, Brunsdon C, 2010. Local forms of spatial analysis. Geographical Analysis, 31(4): 340-358. doi: 10.1111/j.1538-4632.1999.tb00989.x
[16]	Fotheringham A S, Charlton M, Brunsdon C, 1996. The geography of parameter space: An investigation of spatial non-stationarity. Geographical Information Systems, 10(5): 605-627.
[17]	Fotheringham A S, Charlton M E, Brunsdon C, 1998. Geographically weighted regression: A natural evolution of the expansion method for spatial data analysis. Environment and Planning A: Economy and Space, 30(11): 1905-1927. doi: 10.1068/a301905
[18]	Fotheringham A S, Yang W B, Kang W, 2017. Multiscale geographically weighted regression (MGWR). Annals of the American Association of Geographer, 107(6): 1247-1265. doi: 10.1080/24694452.2017.1352480
[19]	Fotheringham A S, Yue H, Li Z, 2019. Examining the influences of air quality in China’s cities using multi scale geographically weighted-regression. Transactions in GIS, 23(1): 1444-1464. doi: 10.1111/tgis.12580
[20]	Geniaux G, Napoléone C, 2008. Semi-parametric tools for spatial hedonic models: An introduction to mixed geographically weighted regression and geoadditive models. Hedonic Methods in Housing Markets, 101-127.
[21]	Golub G H, Loan C F, 1980. An analysis of the total least squares problem. SIAM Journal on Numerical Analysis, 17(6): 883-893. doi: 10.1137/0717073
[22]	Han X, Wang P, Wang J et al., 2020. Evaluation of human-environment system vulnerability for sustainable development in the Liupan mountainous region of Ningxia, China. Environmental Development, 34(5734): 100525. doi: 10.1016/j.envdev.2020.100525
[23]	He B, Li X, Quan X et al., 2015. Estimating the aboveground dry biomass of grass by assimilation of retrieved LAI into a crop growth model. IEEE Journal of Selected Topics in Applied Earth Observations & Remote Sensing, 8(2): 550-561.
[24]	Huang L, Ning J, Zhu P, et al., 2021. The conservation patterns of grassland ecosystem in response to the forage-livestock balance in North China. Journal of Geographical Sciences, 31(4): 518-534. doi: 10.1007/s11442-021-1856-6
[25]	Jelinek A, Zalud L, Jilek T, 2019. Fast total least squares vectorization. Journal of Real-Time Image Processing, 16(2): 1-17. doi: 10.1007/s11554-018-0842-4
[26]	Kashki A, Karami M, Zandi R, et al., 2021. Evaluation of the effect of geographical parameters on the formation of the land surface temperature by applying OLS and GWR, A case study Shiraz City, Iran. Urban Climate, 37(4): 100832. doi: 10.1016/j.uclim.2021.100832
[27]	Kumar S, Lal R, Liu D, et al., 2013. Estimating the spatial distribution of organic carbon density for the soils of Ohio, USA. Journal of Geographical Sciences, 23(2): 280-296. doi: 10.1007/s11442-013-1010-1
[28]	Li N, Li T, Wang Y et al., 2016. Monitoring grassland yield in Qinghai Province based on Landsat8 remote sensing image. Journal of Qinghai University, 34(5): 63-68. (in Chinese)
[29]	Li W, Dong F, Ji Z, 2021. Research on coordination level and influencing factors spatial heterogeneity of China’s urban CO₂ emissions. Sustainable Cities and Society, 75(1): 103323. doi: 10.1016/j.scs.2021.103323
[30]	Li X Z, Han G D, Guo C Y, 2013. Impacts of climate change on dominant pasture growing season in central Inner Mongolia. Acta Ecologica Sinica, 33(13): 4146-4155. (in Chinese) doi: 10.5846/stxb201207301082
[31]	Liu H, Huang B, Gao S, et al., 2021. Impacts of the evolving urban development on intra-urban surface thermal environment: Evidence from 323 Chinese cities. Science of The Total Environment, 771(1): 144810. doi: 10.1016/j.scitotenv.2020.144810
[32]	Liu H X, 2019. Spatio-temporal analysis of grassland productivity and research on remote sensing estimation and prediction of grass yield in Inner Mongolia[D]. Shandong: Shandong University of Science and Technology. (in Chinese)
[33]	Liu J, Huang X, He X, et al., 2018. Estimation of grassland yield and carrying capacity in Qinghai Province based on MODIS data. Pratacultural Science, 35(10): 2520-2529. (in Chinese)
[34]	Liu K, 2012. Study on the mechanism of loess landslide induced by earthquake in Yuanzhou District, Guyuan city[D]. Xi’an: Chang’an University. (in Chinese)
[35]	Luo C, Yu X, Liu C, et al., 2015. Estimation of grass yield in large region on geographically weighted regression model. ISPRS: International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XL-7/W3(7): 9-13.
[36]	Luo L, Wang Z M, Ren C Y, et al., 2010. Models for estimation of grassland production and spatial inversion based on MODIS data in Songnen Plain. Transactions of the Chinese Society of Agricultural Engineering, 26(5): 182-187. (in Chinese)
[37]	Mackey Brendan G, 1994. Predicting the potential distribution of rain-forest structural characteristics. Journal of Vegetation Science, (4): 43-54.
[38]	Mansour S, Al K A, Al Said A, et al., 2020. Sociodemographic determinants of COVID-19 incidence rates in Oman: Geospatial modelling using multiscale geographically weighted regression (MGWR). Sustainable Cities and Society, 65(2): 102627. doi: 10.1016/j.scs.2020.102627
[39]	Marquardt D W, 1970. Generalized inverses, ridge regression, biased linear estimation, and nonlinear estimation. Technometrics, 12(3): 591-612.
[40]	Mei C L, He S Y, Fang K T, 2004. A note on the mixed geographically weighted regression model. Journal of Regional Science, 44(1): 143-157. doi: 10.1111/j.1085-9489.2004.00331.x
[41]	Nakaya T, 2001. Local spatial interaction modelling based on the geographically weighted regression approach. GeoJournal, 53: 347-358. doi: 10.1023/A:1020149315435
[42]	Niu Z, Ni S, 2003. Study on models for monitoring of grassland biomass around Qinghai Lake assisted by remote sensing. Acta Geographica Sinica, 58(5): 695-702. (in Chinese)
[43]	Oshan T M, Li Z, Kang W et al., 2019. MGWR: A python implementation of multiscale geographically weighted regression for investigating process spatial heterogeneity and scale. International Journal of Geo-information, 8(6): 269.
[44]	Qian S, Fu Y, Pan F F, 2010. Climate change trend and grassland vegetation response in growing season in Sanjiangyuan area. Scientia Sinica Terrae, 40(10): 1439-1445. (in Chinese) doi: 10.1360/zd2010-40-10-1439
[45]	Qin N, Ren N, Huang N et al., 2021. Application of geographically weighted regression model in the estimation of surface air temperature lapse rate. Journal of Geographical Sciences, 31(3): 389-402. doi: 10.1007/s11442-021-1849-5
[46]	Qin W, Wang J, 2006. Exploring spatial relationship non-stationary based on GWR and GIS. Proceedings of SPIE: The International Society for Optical Engineering, 6420: 64201X-8.
[47]	Qiu L, Kai W, Long W et al., 2016. A comparative assessment of the influences of human impacts on soil Cd concentrations based on stepwise linear regression, classification and regression tree, and random forest models. Plos One, 11(3): e0151131.
[48]	Qiu X F, Qiu Y P, Zeng Y, 2009. Distributed modeling of monthly mean air temperature of rugged terrain of Chongqing. Advances in Earth Science, 24(6): 621-628. (in Chinese)
[49]	Ren J Z, 1998. Research Methods of Grassland Science. Beijing: China Agricultural Press, 1-16. (in Chinese)
[50]	Roy D P, Borak J S, Devadiga S, et al., 2002. The MODIS land product quality assessment approach. Remote Sensing of Environment, 83(1/2): 62-76. doi: 10.1016/S0034-4257(02)00087-1
[51]	Schwarz G, 1978. Estimating the dimension of a model. The Annals of Statistics, 6(2): 461-464.
[52]	Sun C M, Liu T, Tian T, et al., 2013. Remote sensing estimation and application of grassland NPP based on MODIS data in southern China. Acta Prataculturae Sinica, 22(5): 11-17. (in Chinese)
[53]	Sun R H, Chen L D, 2008. Study on vertical differentiation and digital identification of the landscape in mountain areas. Proceedings of the 5th Chinese Young Ecologists Symposium, 326-331. (in Chinese)
[54]	Tian J, Xiong J N, Zhang Y C, et al., 2021. Quantitative assessment of the effects of climate change and human activities on grassland NPP in Altay Prefecture. Journal of Resources and Ecology, 12(6): 743-756.
[55]	Tucker C J, Vanpraet C L, Sharman M J, et al., 1985. Satellite remote sensing of total herbaceous biomass production in the Senegalese Sahel. Remote Sensing of Environment, 17(3): 233-249. doi: 10.1016/0034-4257(85)90097-5
[56]	Vasilis D, Hauke H, Fabian L, et al., 2018. Predicting methane yield by linear regression models: A validation study for grassland biomass Bioresource Technology. Bioresource Technology, 265: 372-379. doi: 10.1016/j.biortech.2018.06.030
[57]	Wang C, Zhen L, Du B, 2016. Assessment of the impact of China’s Sloping Land Conservation Program on regional development in a typical hilly region of the loess plateau: A case study in Guyuan. Environmental Development, 21: 66-76. doi: 10.1016/j.envdev.2016.11.002
[58]	Wang K B, Li J P, Shangguan Z P, 2012. Biomass components and environmental controls in Ningxia grasslands. Journal of Integrative Agriculture, 11(12): 2079-2087. doi: 10.1016/S2095-3119(12)60466-3
[59]	Wang S J, Fang C L, Ma H T, et al., 2014. Spatial differences and multi-mechanism of carbon footprint based on GWR model in provincial China. Journal of Geographical Sciences, 24(4): 612-630. doi: 10.1007/s11442-014-1109-z
[60]	Wang X, 2008. Study on biomass and species diversity on environmental gradient of alpine grassland[D]. Urumqi: Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences. (in Chinese)
[61]	Wei C H, Qi F, 2012. On the estimation and testing of mixed geographically weighted regression models. Economic Modelling, 29(6): 2615-2620. doi: 10.1016/j.econmod.2012.08.015
[62]	Wei Y X, Wang L W, Shi Y C, et al., 2012. Net primary productivity of grassland resources monitoring based on remote sensing data in Qinghai Province. Scientia Geographica Sinica, 32(5): 621-627. (in Chinese)
[63]	Wright I A, 2010. Field and laboratory methods for grassland and animal production research. Grass & Forage Science, 57(2): 189-189.
[64]	Xie Y, Sha Z, Yu M, et al., 2009. A comparison of two models with Landsat data for estimating above ground grassland biomass in Inner Mongolia, China. Ecological Modelling, 220(15): 1810-1818. doi: 10.1016/j.ecolmodel.2009.04.025
[65]	Xu B, Yang X C, Tao W G, et al., 2007. Remote sensing monitoring upon the grass production in China. Acta Ecologica Sinica, 27(2): 405-413. (in Chinese) doi: 10.1016/S1872-2032(07)60012-2
[66]	Xu X, Xu Y, Sun Q Q, et al., 2018. Comparison study on meteorological spatial interpolation approaches in Kangdian region of China. Journal of Central China Normal University, 52(1): 122-129. (in Chinese)
[67]	Yang S H, Liu F, Song X D, et al., 2019. Mapping topsoil electrical conductivity by a mixed geographically weighted regression kriging: A case study in the Heihe River Basin, Northwest China. Ecological Indicators, 102(C): 252-264.
[68]	Yang Z, Xia J, Zou L, et al., 2021. Efficiency and driving force assessment of an integrated urban water use and wastewater treatment system: Evidence from spatial panel data of the urban agglomeration on the middle reaches of the Yangtze River. Science of The Total Environment, 805(2): 150232. doi: 10.1016/j.scitotenv.2021.150232
[69]	You H Y, Luo C F, Liu Z J, et al., 2014. Study on the method of grass yield model in the Source Region of Three Rivers with multivariate data. IOP Conference Series Earth and Environmental Science, 17(1): 012031.
[70]	Yu H, Fotheringham A S, Li Z, et al., 2019. Inference in multiscale geographically weighted regression. Geographical Analysis, 52(1): 1-20. doi: 10.1111/gean.12206
[71]	Yu H, Fotheringham A S, Li Z, et al., 2020. On the measurement of bias in geographically weighted regression models Spatial Stats. Spatial Statistics, 38: 100453. doi: 10.1016/j.spasta.2020.100453
[72]	Yu H J, Lu X M, Yue X X, et al., 2009. Dynamic change analysis and evaluation of forest resources in Yuanzhou District of Guyuan City. Modern Agricultural Science and Technology, (14): 181-183. (in Chinese)
[73]	Zeng C, Yang L, Zhu A X, 2016. Mapping soil organic matter concentration at different scales using a mixed geographically weighted regression method. Geoderma, 281: 69-82. doi: 10.1016/j.geoderma.2016.06.033
[74]	Zhang Q, 2019. Temporal and spatial patterns and dynamic changes of grass yield in Hulunbeier grassland[D]. Hohhot: Inner Mongolia University. (in Chinese)

Variable	Abbreviation	Unit	Variable description
Intercept	Intercept	g	Intercept term of the model
Average rainfall in July	AJR	mm	Average July rainfall
Relative humidity	RH	%	Percentage of water vapor pressure in the air vs. saturated water vapor pressure at the same temperature
NPP	NPP	gC·m^‒2·a^‒1	Net primary production capacity of vegetation, refers to the total amount of organic matter accumulated by photosynthesis in unit area and unit time of green plants, minus the remaining part after autotrophic respiration
Ratio resident-area index	PRI	-	Proportion of impervious surface in the surface area per unit area: $P R I=\frac{B L U E}{N I R}$, where blue and NIR are the pixel reflectance values of blue and near infrared wavelengths, respectively
Elevation	DEM	m	Altitude of sample point
Daily temperature range	DTR	℃	Difference between maximum and minimum daily temperatures
Distance from gully	DS	m	Distance from sample point to valley
Distance from road	DP	m	Distance from sample point to road
Distance from river	DR	m	Distance from sample point to river
Night light	NL	-	Night light distribution in the study area

Variable	Tolerance	VIF
AJR	0.244	4.092
RH	0.248	4.029
NPP	0.459	2.180
PRI	0.755	1.325
DEM	0.445	2.249
DTR	0.248	4.025
DS	0.792	1.262
DP	0.804	1.243
DR	0.754	1.327
NL	0.877	1.140

Parameter	OLS	GWR	MGWR
Residual sum of squares	959.816	236.960	23.701
-2 log-likelihood	3697.432	3138.206	2558.904
Classic AIC	3731.432	3270.442	2159.397
AICc	3731.660	3273.816	2951.878
BIC/MDL	3831.737	3660.555	2331.959
CV	5863.375	4370.185	-
R²	0.643	0.801	0.891
Adjusted R²	0.642	0.797	0.889

Variable	Coefficient	T-test	Significance (p)
Intercept	366.649	8.482	0.000
AJR	20.505	3.867	0.000
RH	-14.883	-2.593	0.010
NPP	-7.321	-2.305	0.021
PRI	-6.363	-2.580	0.010
DEM	74.194	27.200	0.000
DTR	-40.112	-7.649	0.000
DS	-17.548	-5.498	0.000
DP	3.870	1.490	0.004
DR	9.427	4.379	0.000
NL	-7.397	-1.181	0.004

Variable	Min	Max	Lwr Quartile	Median	Upr Quartile
Intercept		664.366	165.834	373.563	451.646
AJR	-49.684	92.181	-12.087	6.980	35.114
RH	-39.418	144.282	-11.053	8.416	27.694
NPP	-6.232	32.852	10.403	15.679	23.962
PRI	-7.563	2.873	-3.637	-2.698	-1.007
DEM	-13.540	99.800	-3.165	33.316	59.446
DTR	-95.718	54.511	-24.421	-3.836	12.769
DS	-32.859	-0.367	-20.809	-10.443	-4.266
DP	-38.467	41.975	-7.914	0.182	8.727
DR	-56.344	59.046	-18.597	-2.029	18.442
NL	-53.406	56.678	-10.603	-1.480	10.847