A new basis for assessing the onset and withdrawal dates of the Indian monsoon is derived from variability of the large-scale monsoon hydrologic cycle. The hydrologic cycle is chosen as an important diagnostic with which to monitor the monsoon due to the key role played by the zonal and meridional gradients in water vapor, clouds, and rainfall in driving the its large-scale circulation of the monsoon. Among the parameters comprising the index are vertically integrated moisture transport (VIMT) and Convective Available Potential Energy (CAPE). These fields are assessed over large-scale regions where fluctuations are both rapid and large during monsoon onset. The result is the formulation of a hydrologic onset and withdrawal index (HOWI) that provides an objective assessment of monsoon onset and withdrawal dates while being relatively insensitive to synoptic variability. HOWI is thus an alternative to idices that rely on rainfall over a limited domain and subjective assessments of the large-scale monsoon circulation.

Analysis of interannual variability of monsoon onset based on the HOWI reveals robust associations between monsoon onset, total seasonal rainfall, ENSO, and the SOI that are generally feeble and insignificant when assessed using other onset criteria. For example, while the onset and total JJAS Indian rainfall experience anticorrelations weaker than -0.11 when onset date from the Indian Meteorological Department (IMD) or other objective methods are considered, the HOWI onset criterion shows a correlation of 0.36. Additionally the timing of monsoon withdrawal based on HOWI shows strong association with JJAS rainfall correlating at -0.45. Thus the length of the monsoon season is strongly related to overall monsoon strength. Also, though correlation between IMD onset date and Niņo-3 SST is weak (<0.17), monsoon onset based on HOWI is shown to exhibit a modest correlation with Niņo-3 SST at 0.41. Also, monsoon withdrawal based on HOWI correlates at 0.25 with Niņo-3 SST. Thus an important and statistically significant relationship is shown to exist between ENSO and the length of the monsoon season. A variety of applications for the new index are discussed.