*2.5. Topographic Potential Index*

Because the distribution of land-use types is highly related to elevation and slope, this paper adopts the Topographic Position Index composed of slope and elevation to reflect the comprehensive topographic features of the paddy field and dryland.

$$TPI = \lg\left[\left(\frac{E}{E\_0} + 1\right) \* \left(\frac{S}{S\_0} + 1\right)\right] \tag{1}$$

where *TPI* refers to the Topographic Potential Index; *E* and *E*<sup>0</sup> are the elevation and the average elevation of an area, respectively; *S* and *S*<sup>0</sup> are the slope and the average slope of an area, respectively. The *T* value increases along with the increase in slope and elevation. When the slope increases while the elevation decreases, or when the elevation increases while the slope decreases, the value of *TPI* is in the middle of the two. Referring to the relevant literature [23], this paper classified the topographic position index (abbreviated as TPI) into three levels: high (>0.8), medium (0.4–0.8), and low (<0.4).

### **3. Result Analysis**

#### *3.1. Changes in the Topographic Characteristics of Paddy Field and Dryland*

The results showed some similarities in the distribution of the TPI of paddy field and dryland (Figure 1). Firstly, paddy fields and drylands with a low TPI were mainly distributed in the low-elevation areas of eastern China. Specifically, the paddy fields with low a TPI were mainly distributed in the Yangtze Plain areas (including Jiangsu, Shanghai, Anhui, Zhejiang, Hunan, Hubei, and Sichuan), the southern areas of China (Guangdong and Taiwan), the northeastern area (Liaoning and Heilongjiang), and the northeast area (Ningxia Province), while the drylands with a low TPI were mainly distributed in the northern area (Shandong, Henan, Hebei, and Shaanxi provinces), the northeastern area (Liaoning, Jilin, Heilongjiang, and Inner Mongolia provinces), the northwestern area (Xinjiang), and the southern area (Guangdong Province). Secondly, the paddy fields and drylands with a medium TPI were mainly scattered in the second ladder (1000–2000 m elevation) area and small parts of the first ladder (>4000 m elevation) and third ladder (<500 m elevation) in China. Specifically, the paddy fields with a medium TPI were distributed in the south of Shaanxi Province, and the drylands were distributed in Shaanxi, Shanxi, Gansu, Ningxia, Chongqing, Sichuan, Guizhou, Guangxi, Yunnan, and other provinces. Thirdly, the paddy fields and drylands with a high TPI were scattered in the southwest of China, and the drylands area far surpassed that of paddy fields.

An obvious change occurred in the topographic condition of arable land. The paddy fields with a low TPI decreased (Table S1 in the Supplementary Materials), while the paddy fields and drylands with a medium and high TPI increased. Specifically, the proportion of paddy fields with a low TPI decreased from 86.9% to 84.9%, and the proportion of drylands with a low TPI decreased from 76.5% to 75.4%. Because the low TPI areas are mainly distributed in the third ladder of China, which is flat and low in elevation, the hydrothermal conditions and the development degree of its paddy fields and drylands are better, the loss of paddy fields and drylands with a low TPI due to the fact of urban expansion is regrettable, which may lead to a further decline in the suitability of agricultural production [24].

From the perspective of elevation (Figure 2), 40% of the drylands and 64% of paddy fields were distributed at 0–200 m elevation, which decreased over time. The dryland area was greater than that of paddy field at the same elevation. From 1990 to 2020, the most obvious change in dryland and paddy field was concentrated at an elevation of 0–200 m. When looking at the change in dryland, this study found that the dryland increased in all elevations except for 0–200 m, with the highest annual growth rate (1.01%) in the >3500 m elevation area and the lowest annual growth rate (0.03%) in the 1500–2500 m elevation area. When looking at the change in paddy field, the results showed that the paddy field area increased at 200–500 m, 2500–3500 m, and >3500 m elevations but decreased at the other elevations, with the highest average annual growth rate (2.25%) at >3500 m elevations and the lowest average annual growth rate (0.01%) at 200–500 m elevations.

In terms of the slope condition (Figure 3), most of the paddy fields and drylands were concentrated at the 0–2◦ slope and decreased along with the increase in the slopes. From 1990 to 2020, paddy fields and drylands gradually moved from lower to higher slopes; the area of paddy fields and drylands in flatter areas (0–6◦ slope) decreased, whereas the area of paddy fields and drylands in areas with a slope > 6◦ increased. When looking at the change trends, dryland showed a fluctuating change (first increased and then decreased at 0–2◦ and 2–6◦ slopes), while paddy field showed a steadily decreasing trend at 0–2◦ and 2–6◦ slopes. When looking at the obvious change, dryland changed the most at a 6–15◦ slope (increased by 15,026 km2), whereas the paddy field area changed the most at a 0–2◦

slope (decreased by 18,846 km2). When looking at the fastest change, the dryland area and paddy field area increased the fastest (0.6% and 1.09% a year) at a slope of >25◦. In summary, the increase in sloping arable land changed significantly, and the emergence of steeply sloping paddy fields and dryland is especially alarming. Because sloping paddy field and dryland have limits regarding soil and water conservation [17], to guarantee the crop yield sustainably, more slope treatment practices, such as drainage ditches, protection forest, silt arresters, and contour ploughing (terrace field), are needed in this area.

**Figure 1.** Distribution of paddy field and dryland under different TPIs in China in 1990 and 2020. ((**a**), low (<0.4) TPI in 1990; (**b**), medium (0.4–0.8) TPI in 1990; (**c**), high (>0.8) TPI in 1990; (**d**), low (<0.4) TPI in 2020; (**e**), medium (0.4–0.8) TPI in 2020; (**f**), high (>0.8) TPI in 2020).

**Figure 2.** Changes in the elevations of paddy field and dryland. (For detailed data, refer to Table S2 in the Supplementary Materials).

**Figure 3.** Change in the slopes of paddy field and dryland. (For detailed data, refer to Table S3 in the Supplementary Materials).

When looking at the slope aspects (Figure 4), the results showed that the paddy fields and drylands were mainly distributed on the southern, southwestern, southeastern, eastern, and western slopes of hills or mountains. Moreover, important to note is that the area of paddy fields and drylands on southeastern, southern, and southwestern slopes generally decreased, while the area of paddy fields and drylands on northwestern, northern, and northeastern slopes generally increased. Moreover, the area of paddy fields and drylands increased the most in the north, whereas the area of paddy fields decreased most in the south, and drylands decreased most in the southeast. The results indicate that the area of paddy fields and drylands with better light conditions decreased. Under the influence of man-made disturbance in the urbanization process, the advantages of ideal light and heat resources on arable land were ignored, and the area of paddy fields and drylands with superior natural conditions were reduced.

**Figure 4.** Change in the slope aspects of paddy field and dryland. (For detailed data, refer to Table S4 in the Supplementary Materials).
