The five years of summertime days were classified into boundary days (BDs) and nonboundary days (NBDs). BDs were those with at least one identified boundary occurring, whereas NBDs were those without any identified boundaries occurring. A BD was further labeled as a convective boundary day (CBD) or nonconvective boundary day (NCBD), depending on whether or not at least one identified convective boundary appeared. There were 194 (42.2%) BDs out of the total 460 summertime days, with 100 days (21.7%) being CBDs.
The daily atmospheric circulations in these five summers were classified into five distinct synoptic patterns based on an objective classification method (T-mod PCA). Three synoptic patterns (T4, T5, and T1) that contributed to ∼58% of the summer days were found to be favorable for both BD and CBD occurrence. Under the most frequent one of these three patterns, HID was dominated by a high-pressure ridge. Under the other two patterns, HID was located pre-trough (either shallow or deep trough). The two unfavorable patterns were both post-trough with different trough locations. Compared with the unfavorable patterns, the favorable patterns were characterized by a warmer near-surface environment over HID with dominant southerly ambient flows opposite to the irrigation-area breeze of the desert-irrigation circulations, which might have enhanced the circulations and resulted in more frequent BD occurrence. Consistent with the features of the Lake Breeze Index proposed in a previous study, the ratios of the perpendicular 10-m wind speed squared to the 2-m temperature under the three favorable synoptic patterns were found to inversely correlate well with the BD occurrence frequencies in HID with sharp vegetation contrast.
It was also found that the larger CAPE and stronger low-level moisture flux convergence under the favorable patterns may have contributed to the active occurrence of CBDs. Under each synoptic pattern, BDs, and especially CBDs, tended to have a warmer environment and more opposing ambient flows than NBDs. Moreover, CBDs can be successfully distinguished from NCBDs based on their significantly stronger upward flow, larger CAPE, and higher moisture content and K index. Under the favorable patterns, the boundaries tended to show more apparent arid-area dependence in terms of spatial distribution, were more dominantly orientated along the main desert-irrigation interface, more moved northward, formed and obtained their maximum length earlier, and had a more dispersive CI time.
Despite this study having related the behaviors of mesoscale boundaries with synoptic impacts, the finer-scale effects, such as those from the surface sensible/latent heat flux gradient across the area of vegetation contrast, have not been addressed, mainly because of the limitations imposed by the sparseness of observations over HID. A field experiment over this region to fill this knowledge gap is necessary, which has been successfully carried out in the summer of 2022, called the DEsert-oasis COnvergence line and Deep convection Experiment (DECODE). Further research based on DECODE results is underway.

