Due to the influence of human activities, land use cover change occurs sometimes in drylands. For example, people return crops to forests or grass to obtain greater ecological benefits. In order to find the optimal ecological decision, we explored the vegetation conversion potential under AWR constraints. The yellow and cyan color in Fig.\ref{478062} reflects the potential of vegetation that can be converted in water-surplus areas, and the brown part reflects how much more vegetation needs to be converted in deficit areas to keep the water balance. In most areas, the surplus water was sufficient to support the conversion of vegetation (>100%). However, it's worth noting that the additional water in the central drylands was only sufficient for partial vegetation conversion (<50%), as shown in the cyan area of Fig.\ref{478062}. Therefore, vegetation conversion in these areas should be more deliberate.
In water-scarce areas, we found that converting grasslands to crops (Fig.\ref{478062}d, e) or converting irrigated to rain-fed crops (Fig.\ref{478062}f) can compensate for water deficits in parts of the west. However, in most regions, even if all existing vegetation was converted to another water-saving vegetation, it could not make up for the shortage of water resources. The results showed that 158.55% (Fig.\ref{478062}a), 125.20% (Fig.\ref{478062}b), 109.64% (Fig.\ref{478062}c), 48.55% (Fig.\ref{478062}d), 70.20% (Fig.\ref{478062}e), and 91.42% (Fig.\ref{478062}f) of the vegetation must be converted, respectively, to balance the local water resources. This implies that the dominant factor in the scarcity of local water resources may not be vegetation and that other water conservation measures are needed.
4. Discussion
- We used three ET data, three precipitation data, one human water use data to calculate water resources, and three ET data and one GPP data to calculate WUE, obtained the following results. It can be found that WUE causes less uncertainty than water resources in the study of current vegetation potential (Fig.\ref{455510}a). Among the nine data combinations for water resources, precipitation had a greater effect on the results than ET. High precipitation in ERA resulted in higher potential values, and CMFD was the opposite(Fig.\ref{455510}a).