**1. Introduction**

Plant phenology is the time of a certain growth event in the growth cycle, such as germination, branching, leafing, flowering, fruiting, defoliation and dormancy [1–4]. It directly or indirectly regulates several processes of plant growth, such as carbon and water cycle, playing a crucial role in the earth system [5,6]. Adapting to seasonal changes of the environment, plants show a growth rhythm, which is sensitive to environmental change [7,8]. As one of the most critical factors affecting plant phenology, an increase in temperature can promote the activity of enzymes, thereby prolonging plant development. Specifically, an increase in spring temperature promotes the release of plant dormancy in spring, and generally extends the growth cycle of plants [9–14].

Urbanization is an important feature of world development today, and it is one of the main causes of global environmental change in the 21st century. The acceleration of

**Citation:** Ji, Y.; Jin, J.; Zhan, W.; Guo, F.; Yan, T. Quantification of Urban Heat Island-Induced Contribution to Advance in Spring Phenology: A Case Study in Hangzhou, China. *Remote Sens.* **2021**, *13*, 3684. https://doi.org/10.3390/ rs13183684

Academic Editor: Alfredo Huete

Received: 11 August 2021 Accepted: 13 September 2021 Published: 15 September 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

urbanization in recent years has produced substantial impacts on plant phenology over both urban areas and their rural surroundings [15–19]. This is mostly associated to the local warming effect induced by the urban heat island effect, which resulted from the increase in impervious surface percentage and anthropogenic heat emissions [20–23]. Moreover, it is as well as through the fertilization effect induced by the increase in the concentration of carbon dioxide (CO2), nitrogen oxides (NOx), and other atmospheric trace gases over urban surfaces [24–26]. These changes affect urban environments that plants depend on, and have impacts on the growth of plants, thereby changing the plant phenology [27,28].

At present, many studies have paid attention to impacts of urbanization on the change of plant phenology [27–32]. There are two methods to explore the impacts above: the historical comparison method and the urban–rural comparison method. The historical comparison method compared the phenology before and after urbanization, which was mainly for fast-developing cities [31]. However, due to the difficulty of obtaining long time series data, the historical comparison method is greatly restricted. The urban–rural comparison method used the data of the urban and the rural at the same time to explore the impact of urbanization on phenology, which is a method of changing space for time. The second one has been widely used, because of the grea<sup>t</sup> advantages in large-scale observations of remote sensing data [15–19,32]. Meng et al. investigated the urban and rural phenology of the of 85 giant cities in the continental United States from 2001 to 2014, and the results showed that the start of growing season (SOS) in the urban was 6 days earlier than that in the rural [33]. Wohlfahrt et al. found that with the acceleration of urbanization, the SOS advanced and the senescence delayed in the urban areas where the temperature rises [34]. Hu et al. used the Enhanced Vegetation Index (EVI) to explore the spatio-temporal changes of plant phenology and its response to land surface temperature (LST) in Northeast China [35]. The results showed that the LST was significantly negatively correlated with the SOS. Recently, most current studies focused on varieties of plant phenology and influences of temperature on plant phenology under urbanization, but did not quantitatively evaluate the contribution of the temperature differences to the phenological differences between the urban and the rural. That is, the quantitative contribution of the local warming induced by the urban heat island effect (the difference of LST between urban and rural, ΔLST) to the difference of spring phenology (SOS) between urban and rural (ΔSOS) was less understood in past research.

With the development of statistical methods, it was possible to distinguish the influence of different factors. Li et al. used a statistical method to quantify the contribution of cooling and water supply to the yield benefits due to irrigation. They found that 16% of irrigation yield increase was due to irrigation cooling while the rest (84%) is due to water supply and other factors [36]. Besides, Zhao et al. also used a statistical method to quantifying the impacts of urbanization on vegetation growth. They found that the growth enhancement offset about 40% of direct loss of vegetation productivity caused by replacing productive vegetated surfaces with nonproductive impervious surfaces [16,30]. Based on the studies above, a statistical model was used to carry out this study.

Therefore, the objective of this study is to explore impacts of urbanization on SOS and exploratively distinguish contributions of local warming induced by the urban heat island effect (ΔLST) and other factors to the difference of spring phenology between urban and rural (ΔSOS). Hangzhou, a typical subtropical metropolis, was selected as the study area. Specifically, the spatial differences and inter-annual changes of the phenology in the urban and the rural were compared through a gradient analysis method using satellite-based phenology and LST data from 2006 to 2018. Then, the coupling relationship between phenology and temperature were investigated. After that, taking typical forest grid cells in the urban and the rural areas of Hangzhou as test samples, the local SOS was extracted using a remote sensing vegetation index from 2006 to 2018, and the difference of responses of SOS to LST between the urban and rural was explored. Finally, we exploratively distinguish quantitative contributions of the ΔLST and other factors to the ΔSOS.

#### **2. Materials and Methods**
