Explicitly analyze the soil erosion in the karst and 1 non-karst area of different morphological types in 2 Guizhou of China 3

: How to explicitly understanding the soil erosion intensity change in different 7 geomorphological types is one of key issues in the field of soil and water conservation. According 8 to classification criterion of soil erosion intensity of China, the spatial soil erosion data with the 9 resolution of 10 m×10m in Guizhou Province were obtained by combing with the multi-resolution 10 remote sensing data of ALOS, ZY-3, GF-1, Landsat and GDEMV2, and 2762 field sampling data in 11 2010 and 2015, respectively. a spatial analysis model of soil erosion was improved to analyze the 12 spatiotemporal change of soil erosion intensity in karst and non karst area of Guizhou province, 13 which involved the spatial soil erosion data and different geomorphological type data of Guizhou 14 province. The results show that the soil erosion intensity decreased by 6468.13km 2 in Guizhou 15 Province from 2010 to 2015. The dynamic change intensity in the high-altitude area is larger than 16 in the low-altitude area. The soil change intensity in karst area is higher than in non karst area, 17 especially in the high and middle elevation area in Guizhou province. Moreover, the decreasing ratio 18 of soil erosion intensity in karst area is generally larger than in non karst area, which can be used to 19 explain that the ecological restoration projects and water soil conservation polices carried out in 20 karst area has a good effect, especially in western of Guizhou province from 2010 to 2015, one the 21 other hand, the soil erosion in non karst area should also be focused by local government in the 22 future. 23

study (Poesen 2018). There is no C layer between soil and it is very serious that bedrock and the 31 soil erosion caused by denudation, dissolution and cross distribution of denudation and dissolution, 32 especially in the wide distributed area where karst has characteristic of the vertical development (He 33 et al. 2009). 34 Guizhou Province is a typical representative area of karst landforms in Southwest China where 35 the landform pattern is complex ( Yang et al. 2019). The increasing human activities have caused 36 extremely serious negative interference to this area, which makes the soil erosion particularly 37 serious (  the spatial and temporal distribution of soil erosion and its intensity and direction. At present, the 61 research mainly focuses on the driving effect of land use types on soil erosion, as well as soil erosion 62 mechanism or single terrain factor at small watershed scale. However, at the provincial level, 63 especially in karst areas with fragile ecological environment, there is a lack of spatiotemporal 64 variation intensity of soil erosion in different geomorphic types, especially in karst and non karst 65 areas quantitative analysis of driving factors. 66 Since the 21st century, due to the input and implementation of ecological protection projects 67 and soil and water conservation policies in Guizhou Province, soil erosion has undergone significant 68 spatiotemporal differentiation and change. Therefore, the purpose of this paper is to construct the 69 spatio-temporal analysis model of soil erosion dynamic degree, breadth and relative change rate of 70 different geomorphic types by combining the data of Guizhou landform type with the soil erosion 71 intensity data of 10 × 10 m resolution. Based on the quantitative analysis of soil erosion change 72 intensity of different geomorphic types and karst and non karst areas, the spatial and temporal 73 variation intensity difference of soil erosion in karst and non karst areas is revealed, and the driving 74 effects of ecological protection projects and soil and water conservation policies on Soil and water 75 erosion changes are comprehensively analyzed, so as to further implement in Guizhou Province. Combined with the research conception of this paper, the soil erosion intensity changes of karst and 90 non-karst areas in different geomorphic areas are analyzed, and the hills subdivided in karst areas 91 in vector data are appropriately merged, and finally low basin, medium basin and high basin are 92 obtained. There are 12 types of land, low hill, medium hill, high hill, low mountain, low middle 93 mountain, middle mountain, high mountain, low platform and high platform (Fig. 1a). 94 Due to the different sources and scales of soil erosion data, it is necessary to unify the data to 95 form a resolution of 10 × 10 m. The resolution of ZY-3 and GF-1 is resampled for 10 × 10 m, which 96 is consistent with the resolution of ALOS Image. The resolution of vegetation coverage, bedrock 97 exposure rate, slope and other data is increased to 10 × 10 m to ensure the unity of various factors 98 in the process of comprehensive superposition. 99 Considering the great difference of soil erosion characteristics between karst and non-karst 100 areas in Guizhou, such as soil erosion intensity, the year limitation for soil resistant erosion, risk 101 coefficient of soil erosion, etc., the karst areas refer to the technical standard for comprehensive 102 management of soil erosion in karst areas, and the non-karst areas are based on the classification 103 and classification standards of soil erosion. Combined with land use type, vegetation coverage, 104 bedrock exposure rate, slope and other data, based on a large number of field sampling verification, 105 according to the above standards for karst and non-karst areas, the quantitative discrimination and 106 division of soil erosion grade (Table 1, table 2) are carried out respectively, so as to achieve the 10m 107 × 10m resolution soil in different geomorphological types of Guizhou Province in 2010 and 2015 108 Soil erosion spatial information data extraction. In the process of processing and analyzing the basic 109 data, in order to ensure the classification accuracy and quality of the data, a large number of field 110 sampling verification were carried out on the above classification results in the field, with a total 111 number of 2762, and the verification accuracy in 2010 and 2015 were 85.44% and 91.07%, 112 respectively ( Fig. 1b)  index model of soil erosion in Guizhou Province was constructed, which can be expressed as: 129 Where Di is the dynamic index of soil erosion, with a value of 0~1, expressed as a percentage; 131 , ij S  A represents the net area of change between soil erosion of grade i and soil erosion of grade j 132 (km 2 ); Sa is the area of a geomorphic type area (km 2 ). The greater the Di value, the more intense the 133 change in this area; the dynamic degree Di focuses on the process of soil erosion change rather than 134 its results, which aims to reflect the intensity and overall situation of the change. 135 To reveal the main conversion types of soil erosion intensity in different geomorphic types, 136 combined with the spatial distribution data of different geomorphic types, a soil erosion breadth 137 index model which can be used to calculate the main conversion types of soil erosion in different 138 geomorphic types was constructed： 139 Where Ci is the breadth index of soil erosion, the value is 0~1, expressed as a percentage; Sij is 141 the change area from type i soil erosion to type j soil erosion in a certain area from 2010 to 2015 142 (km 2 ); S is the total area of various soil erosion changes in the area (km 2 ). The higher the Ci value, 143 the more dominant the soil erosion from type i to type j in this region. In view of the fact that there 144 are many types of soil erosion conversion among different geomorphic types, in order to analyze 145 the main conversion types of different soil erosion intensities in each geomorphic type area, the 146 cumulative sum of Ci values from large to small is greater than 70%. 147 On the basis of quantitative analysis of soil erosion extent, in order to reveal the regional 148 differences of different soil erosion intensity changes, the relative change rate model of soil erosion 149 was constructed： 150 Where Ki is the relative change rate of soil erosion, the value is 0~1, expressed as a percentage; 152 the larger Ki value in a certain area indicates that the transition from type i soil erosion to type j soil 153 erosion is more likely to occur than that in other regions. 154 3 Results 155

Comparison of soil erosion intensity between karst and non-karst area 156
The comparative analysis of different levels of soil erosion intensity in karst and non-karst 157 areas of Guizhou Province shows that: from 2010 to 2015, the overall erosion intensity in Guizhou 158 Province showed a weakening trend, with the proportion of erosion area except slight erosion in the 159 total area of Guizhou Province decreased from 31.37% in 2010 to 27.70% in 2015, and the total 160 weakened area was 6468.13 km 2 (Table 3). Among them, the areas of mild, moderate, strong and 161 severe erosion in karst area are higher than those in non-karst area, which are 566.88 km 2 , 794.82 162 km 2 , 265.4 km 2 and 15.17 km 2 more than those in non-karst area. In addition, although the areas of 163 extremely strong and severe erosion have weakened, the weakening extent is the smallest. The 164 weakened areas of soil erosion in karst area and non-karst area only account for 11.03% and 22.42% 165 of the total area of soil erosion weakening in karst area and non-karst area respectively. On the 166 whole, the weakening area of soil erosion intensity in Guizhou karst area is larger than that in non-167 karst area, but the weakening trend of extremely strong and severe erosion types is contrary to the 168 general trend of soil erosion intensity weakening (the weakening range of karst area is 11.04% and 169 4.38% lower than that of non-karst area, respectively

Comparison of soil erosion intensity changes in different geomorphic types 176
3.2.1 Regional differences of different soil erosion intensity 177 The dynamic degree (Di) simulation results of soil erosion intensity change in different 178 geomorphic morphology areas of Guizhou Province from 2010 to 2015 shows that with the rising 179 of altitude, the Di value of each geomorphic form area increases continuously, but the Di value of 180 karst area is smaller than that of non-karst area. Therefore, Di value in high altitude areas such as 181 high basin, high hill and high and middle mountain is higher than that in low altitude areas such as 182 low basin, low hill and low mountain. The Di value of karst area is lower than that of non-karst area, 183 especially in high basin, with a difference of 9.35% (Fig. 2b). The results show that the dynamic 184 degree of soil erosion in high altitude area is higher than that in low altitude area, and the dynamic 185 degree of soil erosion in karst area is lower than that in non-karst area.

Distribution characteristics of soil erosion intensity in different geomorphic types 214
Due to the large proportion of low and medium mountains in the whole province, the proportion 215 of basins and platforms is very small. Considering the distribution of the change of soil erosion 216 intensity only from the perspective of area, almost all the areas where the degree of soil erosion is 217 enhanced or weakened are distributed in low and middle mountains. Therefore, in order to further 218 reveal the distribution of soil erosion intensity in different geomorphic types, the relative change 219 rate was used for analysis and comparison. The results show that (Fig. 3), the distribution of 220 decreased soil erosion intensity in karst area is significantly different from that in non-karst area, 221 but the regional difference of enhanced soil erosion intensity is not obvious. Among them, the micro, 222 mild and extremely strong erosion in karst area is further enhanced, and the relative change rate is 223 higher in middle mountain and high mountain area, while in non-karst area, the relative change rate 224 of moderate and strong erosion aggravation in high basin is higher, and the difference between karst 225 area and non-karst area is 0.19% and 0.2%; in karst area, severe, extremely strong and strong erosion 226 is weakened in middle mountain, high and middle mountain and other areas. The relative change 227 rates of high and middle mountains are high, which are 0.63%, 0.66% and 1.15%, respectively. The 228 relative change rates of moderate and mild erosion are higher in high hills and middle basins 229 respectively; in non-karst areas, the relative change rates of severe, extremely strong, strong and 230 moderate erosion are mainly in high basins, with an average of 2.59%; in addition, slight erosion in 231 karst areas and non-karst areas turns to slight erosion in medium The relative change rate of basin 232 and mid hill is high, with an average of 3.19% and 3.0%, and this area is the rapid expansion area 233 of urban construction land. Province has become better, showing a dynamic change pattern of soil erosion in the western part 259 of Guizhou Province, which changes violently in the west and relatively stable in the east (Fig. 2a). 260 On the one hand, due to the main distribution of forest land and cultivated land in karst area, under 261 the drive of ecological and environmental protection policies, slope cultivated land is gradually 262 reduced, so that the breadth of soil erosion in karst area is higher than that in non-karst area, and the 263 conversion is mostly to weak first-order erosion. On the other hand, the severe, extremely strong 264 and intense soil erosion in karst area is mainly concentrated in the middle and high mountains areas, 265 and the relative change rate of soil erosion in these areas is relatively high, which reflects that the 266 ecological restoration project has achieved good ecological effect in karst area. In addition, due to 267 the rapid development of urbanization in the central basin and hilly areas along the Zunyi-Guiyang-268 Anshun Due to the difficulty in obtaining the refined spatial information of soil erosion, the spatial-276 temporal variation trend and regional difference of soil erosion intensity in karst and non-karst areas 277 of Guizhou Province from 2010 to 2015 are only revealed at the resolution of 10m × 10m. In the 278 further work, we will continue to obtain more refined spatial information data of soil erosion, and 279 then realize the dynamic detection and fine identification of spatial and temporal variation patterns 280 and regional differences of soil erosion intensity in Karst and non-karst areas of Guizhou Province, 281 and further implement soil and water loss prevention measures and control measures in karst and 282 non-karst areas in Guizhou Province Rocky desertification ecological restoration project provides 283 dynamic and refined spatial data and method support. 284

Conclusions 285
In view of the lack of quantitative analysis of soil erosion intensity in different geomorphic 286 types, this paper takes Guizhou Province as an example to study the spatiotemporal variation 287 intensity of soil erosion and its driving factors in karst and non-karst regions of different geomorphic 288 types. The spatial and temporal variation intensity of soil erosion in karst and non-karst regions was 289 revealed by constructing spatiotemporal analysis models of soil erosion in different geomorphic 290 types. This study can provide method support and data support for the refined study of soil erosion 291 change intensity in complex geomorphic type area. 292 The results showed that: 1) from 2010 to 2015, the soil erosion intensity in Guizhou Province 293 showed a decreasing trend, with a total area of 6468.13 km 2 . The total area of soil erosion intensity 294 weakening in karst area is larger than that in non-karst area, but the weakening extent of soil erosion 295 intensity of extremely strong and severe erosion in karst area is less than that in non-karst area, 296 which indicates that the soil erosion restoration difficulty of extremely strong and severe erosion 297 area in karst area is higher than that in non-karst area; 2) with the increase of altitude, the dynamics 298 of various geomorphic types are also discussed. The average Di value increased from 6% in low 299 altitude area to 13.61% in high altitude area, and the increase of Di value in karst area was less than 300 that in non-karst area; 3) the main conversion type of soil erosion in each geomorphic form type 301 area was from mild to slight erosion, and the conversion breadth index of soil erosion in karst area 302 was higher than that in non-karst area, although Ci value in low altitude area was only 1.77%; 4) 303 The results showed that the average effect of the project was 10.4% higher than that of the other 304 areas in Guizhou Province. 305