Role of antecedent land surface conditions on North American monsoon rainfall variability

Possible links between North American Monsoon System (NAMS) seasonal [June-July-August-September (JJAS)] precipitation and premonsoon seasonal land surface conditions including precipitation (P), surface air temperature (Ts), soil moisture (Sm), and snow water equivalent (SWE) anomalies are explored during the 1950-2000 period. A statistically significant inverse relationship is found between monsoon precipitation in an area defined as the Monsoon West (Arizona and western New Mexico) and antecedent winter precipitation in the southwestern (SW) United States and the mountainous region in Utah and Colorado (the predictor area). This linkage is strong during 1965-90 and weak otherwise, as has been suggested by previous studies. A land surface feedback hypothesis is proposed to explain this relationship: more winter P leads to more winter and early spring SWE in the predictor area, hence more spring and early summer Sm, and lower spring and early summer Ts, which induces a weaker onset (and less precipitation) of the NAMS and vice versa. All three links in this hypothesis were tested and the existence of a land memory associated with winter precipitation and snow, which can persist until June, was confirmed. However, the results show that this land memory contributes little to the magnitude of NAM precipitation. Winter snow is negatively correlated to late spring Ts in the SW mountainous region, but not in extreme years. In fact, the premonsoon (June) Ts over the U.S. southwest is inversely related to monsoon precipitation, which is the reverse of what is expected based on the hypothesis. The lack of a significant Sm-Ts-P relationship in most of the SW suggests, based on the constructed Sm dataset, that local premonsoon soil wetness conditions play a minor role in the strength of the monsoon. A strong positive relationship between June Ts anomalies and the large-scale midtropospheric circulation before the onset of the monsoon was found, suggesting that the controlling factor for the premonsoon Ts anomalies may not be local (i.e., not from the land surface). The results suggest that further research is needed to elucidate the nature of land-sea-atmosphere interactions as related to the onset of the monsoon.