LAUSR

Goldsmith and Xu (2020) commented on our study of reconstructing the lake level changes of Dali Lake in Inner Mongolia (Jiang et al., 2020). They pointed out that Jiang et al.'s result of between 10 and 7 ka differs from the previous reconstruction (Goldsmith et al., 2017) and attributed the cause of this discrepancy to the “out of stratigraphic order” of three beach ridges; namely, R2 (6.8 ka), R4 (9.8 ka) and R6 (7.7 ka). The law of superposition is the principle used to judge stratigraphic sequence. However, there is no superimposed relationship among these beach ridges, which are physically separated. As stated in Goldsmith et al.'s comments, beach ridges can form during either regression or transgression. Hence, a beach ridge perched at a higher elevation may not be older. A mechanism of preservation of submerged beach ridges has been proposed and described in detail (Jiang et al., 2020). Goldsmith and Xu (2020) claimed that this mechanism could not be valid, although they have not given any argument. Here, we summarise this mechanism again. During the transgression, low beach ridges will be submerged and buried by finer sediments (the deeper the water, the weaker the bidirectional flow is). This overlying layer is relatively thin in comparison to the beach ridge. When the lake level drops, the fine sediments may be easily eroded away by the surf, and the coarse beach ridge sediments will remain in situ if the recurring surf is not too strong. Hence, a buried beach ridge, such as R7 (Jiang et al., 2020) or DL 24_1 (Goldsmith et al., 2017), can be exposed again, without violating any geomorphological principle. It is a natural result that some beach ridges are preserved under alternating old and young deposits. The determination of a “continuous beach ridge sequence” that formed during a regression can only be made by examining the sedimentary age of each beach ridge. However, Goldsmith and Xu (2020) have made this judgement based only on satellite images and a schematic section. This practice is beyond our understanding of geomorphological principles. Goldsmith and Xu (2020) also proposed an alternative explanation for “out of stratigraphic order”: “carving of young beach ridges into sediments that are actually reworked lake sediments from previous high stands”. However, we think this explanation is unreasonable. Optically stimulated luminescence is a dating technique used to date the last time that quartz sediment was exposed to light. The reworked sediments will record the carving time of the young beach ridge. In contrast, reworking will certainly influence radiocarbon dating, which is widely used by Goldsmith et al. (2017). This uncertainty should be tested by additional dating methods. In addition to reworking, unsuitable “lake sediments” and carbon reservoir effects should also be considered as possible causes for the discrepancy between 10 and 7 ka. Laminated silty clay is regarded as a lake sediment (Goldsmith et al., 2017), but laminated silty clay can also form in other low‐energy water bodies, such as swamps and floodplains, which are widely distributed across a dry lakebed. Some lake sediments were deposited by abandoned small lakes situated at high altitudes. It is clear that laminated silty clay originating from nonlake and abandoned small lake environments cannot indicate the lake level of Dali Lake. Goldsmith et al. (2017) did not clarify how to identify the real lake sediments of Dali Lake. Goldsmith et al. (2017) carried out radiocarbon measurements on modern lake and river water dissolved in organic carbon and a submerged aquatic plant (algae). They considered the lake water to be currently at radiocarbon equilibrium with the atmosphere, obviating the need for a reservoir correction. However, the reservoir effect has been reported in work on many lakes in northwest China (e.g. Long et al., 2011; Wu et al., 2011; Zhai et al., 2011). The organic matter in the surface sediments of Dali Lake yielded a C age of 472 yr, and this anomalously old age was considered to be due to carbon reservoir effects (Fan et al., 2017). The modern lake water exhibiting radiocarbon equilibrium with the atmosphere should be ascribed to the reduced river discharge. The river waters do contain aged carbon, and the mean apparent age is 1.5 ka (Goldsmith et al., 2017). Human activities have greatly reduced the river runoff in this catchment. As the main river, the discharge of the Gonggeer River was 0.457×10m in 1961 and decreased to 0.171×10m in 1972 (Tao, 1984). A continuous decrease in discharge since the 1970s can also be inferred from the increasing salinity of Dali Lake (He et al., 1996). Hence, aged carbon supplied to Dali Lake has largely been reduced over the past several decades. This is a possible reason why the modern lake water shows no obvious reservoir effect. The sediment deposited several decades ago should still be corrected to account for the reservoir effect. However, a reservoir correction was not made by Goldsmith et al. (2017).