Monitoring and Assessment on the UNESCO Endangered Heritage Sites using Space Technology-A Case Study on East Rennell, Solomon Islands

： Space technology is an effective approach to monitor the status of World 6 Natural Heritage sites. East Rennell Island has been inscribed on the List of World 7 Heritage in danger since 2013. The site includes approximately 37,000 hectare (ha) and 8 a marine area extending three nautical miles to sea. Deforestation and natural 9 disasters have become the increasing factors threatening its sustainable development. 10 Based on the analyses of multi-source long-time series remote sensing data like MODIS 11 and Worldview data, the forest cover change and its future trends in Rennell Island 12 from 2000 to 2020 have been mapped and assessed using Sen+Mann-Kendall and 13 Hurst index models. A land cover classification system derived from the high-resolution 14 Worldview images was developed as a baseline product for monitoring and analyzing 15 future forest cover changes in East Rennell Island. Our results show that: (1) Area of 16 the vegetation degradation is basically same as that of vegetation improvement from 17 2000 to 2020. (2) The trend of forest cover change is weak in continuity, and significant 18 improvements in damaged vegetation can be implemented but it needs enough 19 protection measures and financial input. (3) This heritage site has a strong ability to 20 regenerate vegetation, and it is recommended to restrict the human activities like the 21 mining, logging, and road construction, which could greatly disturb the unique

disasters have become the increasing factors threatening its sustainable development. 10 Based on the analyses of multi-source long-time series remote sensing data like MODIS 11 and Worldview data, the forest cover change and its future trends in Rennell Island 12 from 2000 to 2020 have been mapped and assessed using Sen+Mann-Kendall and 13 Hurst index models. A land cover classification system derived from the high-resolution 14 Worldview images was developed as a baseline product for monitoring and analyzing 15 future forest cover changes in East Rennell Island. Our results show that: (1) Area of 16 the vegetation degradation is basically same as that of vegetation improvement from 1. Introduction 26 Space technologies represented by 3S technology has the characteristics of real-27 time and high-precision positioning, dynamic multi-temporal resolution data 28 acquisition, intelligent and efficient data management, and analysis for the world 29 heritage sites [1] [1][2][3]. It is urgent to monitor them continuously and put forward Heritage Committee determined that logging is threatening East Rennell's outstanding 56 universal value and has asked the national authorities to provide an impact assessment 57 study of the logging, although it is taking place outside the core area of heritage site.  The wildlife includes 11 species of bat (one endemic) and 43 species of breeding 70 land and water birds (four species and nine endemic subspecies, respectively). The 71 invertebrate life is also rich with 27 species of land snail (seven endemics) and 72 approximately 730 insect species, many of which are endemic. The flora of Lake 73 Tengano is dominated by more than 300 species of diatoms and algae, some of which 74 are endemic. There is also an endemic sea snake in the lake [8]. Earth satellite observations, we use remote sensing data to monitor the forest cover 81 change and classify the land cover that is an effective assessment tool to protect Rennell 82 Island.

84
Rennell Island, as a tropical island, is often obscured by dense clouds, which have 85 a significant impact on vegetation monitoring and land cover classification from optical 86 remote sensing data. By collecting and compiling multi-source remote sensing data and 87 selecting high quality data without or with few clouds, we obtained 432 scenes of

96
Before data analysis, the original data is preprocessed (Fig. 2).  Sen-Trend Calculation Formula: The Where j x is the j th data value of the time series, n is the number of data samples, 140 and Sgn is a step function calculated by the formula (5).

141
Z values can calculate by formula (6): is the variance of S; n is the number of data 143 in the sequence.
Under the 95% confidence level test, the threshold value of Z is 1.96, and the 145 combination of Z value and Sen slope is shown in Table 3.
146 For any positive integer m. The calculation procedure for the Hurst method is as follows:

155
(1) Differential Sequence (2) Mean Sequence There are three forms of the Hurst exponent. The smaller the H value, the weaker 161 the continuity, and the larger the continuity.

166
The Sen+Mann-Kendall analysis method was selected to obtain the distribution 167 results for the changing trend of NDVI in Rennell Island from 2000 to 2020 (Fig. 3) The overall change of East Rennell Island is slightly improved. The degenerated area 180 is largely distributing to Lake Tengano and the inner periphery of the lake, which is 181 more evident and continuous, mainly along some facilities and constructed structures 182 along the lake.  187 We used the Hurst exponent formula to obtain the distribution of persistent 188 characteristics reflecting NDVI changes in Rennell Island from 2000 to 2020 (Fig. 4). in most areas of Rennell Island will be unsustainable if the previous trend continues. 203 The areas with persistent improvement and anti-continuity decrease of NDVI account 204 for 6.5% and 43% of the total area, respectively, and are mainly located in the   with a spatial resolution of 30m were used to extract and calculate the spatial 221 distribution of forest/vegetation cover change (Fig. 6) and change (Fig. 7) in Rennell

222
Island from 2012 to 2018. The red, yellow, and dark green patches in Figure 6 show the  forest area accounts for 52%, considering that Lake Tengano accounts for 44.73% of 256 the total area. Moreover, the original forest area accounts for more than 95% of the land 257 area. As a result, 93% of the land area is primitive forest (see Table 6).   The area of coverage of forests within ERWHS is over 95% of the land area (i.e., not 284 counting the area covered by Lake Tengano, which is nearly 45% of the ERWHS).

285
However, the vulnerability of the study area will increase during the mining and logging 286 activities.

287
(2)The destruction of forests on the island is mainly due to the activities of  Heritage Committee.

305
According to the data analysis, we give the several suggestions as follows:

306
(1) A land cover classification comprising of 10 categories has been established (see Availability of data and materials 383 The authors confirm that data supporting the finding of this study are available.   Flow chart for processing of the preprocessed data Figure 3 NDVI Trend of Rennell Island from 2000 to 2020 Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.  Flow chart for processing of the preprocessed data Figure 3 NDVI Trend of Rennell Island from 2000 to 2020 Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.

Figure 4
Hurst exponential spatial distribution of Rennell Island from 2000 to 2020 Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.

Figure 5
Continuity distribution of vegetation change in Rennell Island from 2000 to 2020 Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors. Changes in forest cover from 2012 to 2018 Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.

Figure 7
Worldview image covering the Rennell Island Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.

Figure 8
Categories of the land cover classi cation system Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.