Urban vegetation phenology analysis and the response to the temperature change *
1
Feng Li1 , Guo Song1, Zhu Liujun2, Fang Xiuqin1, Zhou Yanan1 Department of Geographic Information Science, Hohai University, Nanjing, 211100, China 2 Department of Civil Engineering, Monash University, 3800, Australia * Corresponding author:
[email protected] Abstract—Constructing the high spatio-temporal NDVI time series data is necessary to study the urban vegetation phenology, and the response relationship between vegetation phenology and temperature changes can better understand the urban ecological conditions. In this paper, taking the typical area in Nanjing City as study area, urban vegetation phenology is deriving from HJ NDVI time series. Two ways of the "season" and "season+ trend" are used to extract the vegetation phenology and the relationship of temperature change and the way of “season + trend” is analyzed. The results showed that HJ NDVI time series can be used to research on urban vegetation phenology. Two ways of the "season" and "season+ trend" are similar in the phenology derivation if there is no break; otherwise, they are quite different if the breakpoints exist. And there is an obvious lag of vegetation phenology caused by temperature. Keywords—HJ NDVI time series; urban vegetation; phenological features; temperature change response; Nanjing City
I.
INTRODUCTION
Vegetation is of particular interest as it presents a versatile resource for effectively managing and moderating a variety of problems associated with urbanization. The spatial distribution and abundance of urban vegetation are recognized as a key factor influencing numerous biophysical processes of the urban environment [1,2]. As the understanding of ecosystem services is evolving, researchers are becoming increasingly aware of the importance of urban vegetation. In addition, vegetation phenology is an indicator of climate change, and to some extent, the climate change has a lag to vegetation phenology[3]. NDVI time series data acquired by satellite sensors can reflect terrestrial vegetation growth status, seasonal aspect, and inter-annual variation accurately. It has been widely used in global and regional ecological environment variables monitoring and simulation, dynamic changes of vegetation cover research, vegetation phenology feature recognition and information extraction, and many other fields [1, 2]. The most used NDVI time series is from NOAA AVHRR, SPOT VEGETATION and TERRA / AQUA MODIS and other sensors [4-8]. However, limited by their relatively low spatial resolution, these NDVI time series products can no longer satisfy the fine scale researches especially for urban research. Therefore, constructing the high spatio-temporal resolution NDVI time series and analyzing the relationship between urban vegetation phenology and temperature change are worth to deeply research. The satellite of HJ-1 A/B was launched by China in 2008. Since the launch date, the remote sensing images of HJ have constituted nearly eight years of time-series data making it
978-1-5090-4951-6/17/$31.00 ©2017 IEEE
possible for urban application research due to relative high spatial and temporal resolution. Taking the typical vegetation region in Nanjing City as the study area, urban vegetation phenology is deriving from HJ NDVI time series. Two ways of the "season" and "season+ trend" are analyzed and the S-G filter fitting is used to extract the vegetation phenology of different kinds of urban vegetation. Finally, the relationship between urban phenology and temperature change is analyzed. II.
RESEARCH AREA AND DATA SOURCE
A. Research Area The urban area of Nanjing City, Jiangsu Province (Fig. 1) is selected as the case study area. Nanjing, is located in the largest economic zones of China, the Yangtze River Delta, as part of the downstream Yangtze River. It's extremely scorching in summer and frigid in winter, and the temperature gap turns out to be wide each year. Summer witnesses the largest precipitation. Due to the specific ecosystem, vegetation phenology is significant in this area. And the study area is characterized by a great variety of plants and tree species, including high trees, short-cut trees, shrubs and grassland which means that the several phenological patterns occur in this area.
Fig.1. The location of study area
B. HJ-1A/B images, field data and data pre-processing 77 HJ-A/B images with the uniform interval of 5 days are selected for this research. These images were acquired on the whole of year 2013 which cover the study area with the size of 771 × 391 pixels. The images have a spatial resolution of 30 m
5743
IGARSS 2017
and a breadth of 700 km. Radiometric calibration and atmospheric correction are applied to the images firstly, then, taking HJ satellite CCD image of May 1, 2013 as a benchmark, relative registration is carried out and registration error is controlled in 0.5 pixels. To better understand the vegetation in the study area and collect some pure vegetation samples for the evaluation of filters’ performance a field survey was conducted on April 2014. Specifically, typical urban vegetation samples in Nanjing, i.e. shrub, grassland, evergreen coniferous tree, broadleaved deciduous tree, and evergreen and deciduous broadleaved mixed tree, are collected. The size of each sample is larger than 90 m×90 m to avoid the possible error in NDVI time series caused by the pixel offset in the process of remote sensing image registration. The detailed information of a field survey is shown in Fig. 2.
Table 1 Sample information of pure vegetation points No. 1
Shrub
2
Grassland
3 4 5
Location 118.95E 32.12N 118.90E 32.10N 118.85E 32.03N
Evergreen Coniferous Tree Deciduous Broadleaved Tree Mixed Tree
III.
Quality Level 4 5 1
Remarks NanJing Univ. Xianlin Campus Nanjing Normal Univ. Xianlin Campus Nanjing Univ. of Science and Technology
118.76E 32.06N
3
Southeast Univ. Gulou Campus, Baima Park
118.76E 32.05N
2
Qingliang Mountain Park
The time series data of different kinds of vegetation and temperature will be decomposed. Firstly, not only analyze the changes of two parts of "season", "trend", but also the part of "season + trend", the correlation relationship between the part of "season + trend" and temperature change will be obtained. At the same time, the inter-annual variation of phenological characteristics is related to the temperature in a certain period of pre-time (such as autumn, winter, etc.). By comparing the difference of the mean temperature in different years, it can be used to analyze the change of the start season of vegetation in the whole region. IV.
RESULTS
A. Time series data decomposition It can be seen from the time series decomposition chart of the temperature that change of each part of the daily maximum temperature and the daily average temperature is roughly the same, in the trend part, it can be seen the temperature dropped in the 2010-2012 about 1-2℃. For the change of key phenological points of vegetation, it is necessary to compare the temperature of 2-3 months before the phenological points. From the decomposition charts of NDVI time series data tree five typical vegetation (Fig.3), it can be seen that in the trend part, there is no break points of grassland and deciduous broad-leaved forest and shrub, and 1 break point of mixed tree and has 3 break points of evergreen coniferous tree; in the seasonal part: the curve of grassland has the single peak which is smooth, the remaining four vegetation have the low ebb between June and July caused by the high temperature after the rainy season. The effects of high temperature on shrubs are more severe, and the effects on mixed tree and coniferous forest tree relatively small. This research only focuses on the two parts of season and trend, the remainder is not discussed.
Fig. 2. The route of field survey and the sample points
Vegetation Type
B. The response relation analysis of urban vegetation phenology and temperature change
METHODOLOGY
A. Time series data decomposition BFAST [9, 10] integrates the iterative decomposition of time series into the seasonal, trend and re-remainder components for detecting the phenological changes. The seasonal component can be used to monitor inter-annual phenological changes while the trend can detect the abrupt and gradual change at a longer time scale. Fig.3 Decomposition charts of NDVI time series of five typical vegetation and temperature
5744
B. Phenological features of typical urban vegetation Fig.4 illustrates the break points exist in mixed tree and evergreen coniferous tree, therefore, the great difference is showed between "season" and "season + trend" by comparing the fitting curves of "season" and "season + trend".
From the Fig.5, the correlation between the daily air temperature, the daily maximum temperature and grassland is the best which is the poor for evergreen coniferous tree. It can be found that when the temperature is too high, the growth of vegetation will be weakened, which explains the "trough" phenomenon in the NDVI curves of typical vegetation. In the time series analysis of temperature shown in Fig.3, we can see that the overall trend of temperature in three years has declined. The changes of the start point and the end point of growth season in the whole study area are analyzed by using temperature in autumn and winter. Compared with the winter in 2010, the temperatures of the winter in 2011 are relatively stable, and the mean values of temperature in 2011 are lower than 2010 with 0.4 ℃ which cause a delay of about 20 days in the growing season in 2012. Compared with autumn in 2011, the temperature fluctuations can be shown in autumn of 2012, and the mean values of temperature in 2012 are lower than in 2011 with 1.3 ℃ which has little effect on the overall phenology in the research area. However, for Zijin Mountain scenic area, the end point of vegetation growing season is still delayed, which is associated with the factors of the duration of sunshine and precipitation. V.
CONCLUSIONS
We are trying to research on urban vegetation phenology derived from the high spatio-temporal NDVI time series, and discuss the relationship between urban vegetation phenology and temperature change. And there is an obvious lag of urban vegetation phenology caused by temperature.
Acknowledgment This research was supported by “the National Natural Science Foundation of China (No. 41301446)” and “the Fundamental Research Funds for the Central Universities (2015B11314)”. The authors would like to thank all the organizations for sharing data.
Fig.4 NDVI time series of five typical vegetation (left is “season”, right is “season + trend”)
References
C. Correlation analysis of temperature change and vegetation phenology
Fig.5 Correlation analysis of temperature and vegetation phenology (Season + Trend)
[1] S. A. Ackerman, K. I. Strabala, W.P. Menael, et al. Discriminating Clear Sky from Clouds with MODIS. Journal of Geophysical Research, 1998, 103(32):141-157. [2] F. Yu, K. P. Price, J. Ellis, et al. Satellite Observations of the Seasonal Vegetation Growth in Central Asia: 19821990. Photogrammetric Engineering & Remote Sensing, 2004, 70(4): 461-469. [3] Piao S L, Fang J Y, Zhou L M, et al. Variation in satellitederived phenology in China’s temperate vegetation. Global Change Biology, 2006, 12:672-685. [4] R. Fensholt, K. Rasmussen, T. T. Nielsen, & C. Mbow. Evaluation of earth observation based long term vegetation trends - Intercomparing NDVI time series trend analysis consistency of Sahel from AVHRR GIMMS, Terra MODIS and SPOT VGT data. Remote Sensing of Environment, 113, 2009, pp.1886-1898. [5] J. Chen, P. Jӧnsson, T. Masayuki, et al. A Simple Method for Reconstructing a High-quality NDVI Time-series Data
5745
Set Based on the Savitzky-Golay Filter. Remote Sensing of Environment, 2004, 91:332-344. [6] P. Jӧnsson, L. Eklundh. Seasonality extraction by function fitting to time-series of satellite sensor data. IEEE Transactions on Geoscience and Remote Sensing, 2002, 40 (8), pp.1824-1832. [7] P. Beck, C. Atzberger, K.A. Høgdac et al. Improved Monitoring of Vegetation Dynamics at Very High Latitudes: A New Method Using MODIS NDVI. Remote Sensing of Environment, 2006, 100, 321-334. [8] Z. H. Xue, P. J. Du and L. Feng. Phenology driven land cover classification and trend analysis based on long-term remote sensing image series. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2014, 7 (4):1142 -1156. [9] Verbesselt J, Hyndman R, Zeileis A, et al. Phenological change detection while accounting for abrupt and gradual trends in satellite image time series. Remote Sensing of Environment, 2010, 114(12): 2970-2980. [10]Verbesselt J, Zeileis A, and Herold M. Near real-time disturbance detection using satellite image time series. Remote Sensing of Environment, 2012, 123(123):98-108.
5746
