This dataset was used for the research described in the article Machine Learning Predictions of Deep Convection Initiation: The Importance of Spatial Information, the Sensitivity to Training Region, and Dominant Environmental Parameters published in the American Meteorological Society's Artificial Intelligence for the Earth Systems. Environmental parameters are derived using vertical profiles from Rapid Update Cycle (RUC) or Rapid Refresh (RAP) model analyses. Data key Point_Type: "Null" or "DCI" Latitude: Degrees north Longitude: Degrees east Time(NearestHourUTC): MM/DD/YYYY HH:mm MU_CAPE: Most-unstable (MU) convective available potential energy (CAPE) MU_CIN: MU convective inhibition MU_Max_OMEGA: MU maximum value of vertical motion (omega) below 100 hPa above the MU level of free convection (MU-LFC) MU_OMEGA_Height: The height at which the maximum value of vertical motion below 100 hPa above the MU-LFC occurs MU_HLFC: The ratio of the height of the maximum omega to the height of the MU-LFC MU_Mean_Convergence: Convergence for the layer from MU parcel height to MU-LFC MU_Subcloud_Shear: Vertical wind shear in the layer from MU parcel height to the lifting condensation level (LCL) MU_ACBL_Shear: Vertical wind shear in the 1.5 km layer above the MU-LFC MU_ACBL_Lapserate: The laspse rate of the active cloud-bearing layer defined as the 1.5 km deep layer above the MU-LFC MU_CD: The height between the MU parcel and and the MU-LFC MU_LCL2CAPE: CAPE of the MU parcel in the 2km layer above the LCL MU_MRD: The intergrated difference in mixing ratio between the parcel and the environment from the MU height to the MU-LFC