Water Vapor Trajectory Forecasting
Two- and three-day forecasts of upper tropical troposphere/lower stratopsheric water vapor (also CO and O3) are
generated once daily for the 375 K level using the Reverse Domain-Filling (RDF) approach. Specifically, back
trajectories at 375 K are initialized on a 0.5 x 1.0 degree latitude-longitude grid, with the initial water
vapor, CO and O3 fields derived from an interpolation at Day 0 from 2 days of MLS near-real time (NRT) water
vapor, CO and O3. Trajectory winds and temperature are from the GFS forecasts; an offset of -2 K is heuristically
applied to temperature to account for uncertainties in the supersaturation as well as temperature bias.
Once the MLS water vapor is initialized on the trajectory, the water is dehydrated along the trajectory path assuming
clouds are forming in saturated air. Pfister's convective influence is included if the trajectory passes through
a convective system that reaches the the parcel level (minus 0.5 km or above). If convective influence is encountered,
the parcel is reset to the local saturation mixing ratio.
Explanation of plots provided
a. MLS H2O at 375 K: MLS NRT data for days -1 and 0 interpolated to a global grid
b. Temperature plus flow at 375 K
c. Cloud probability: For these calculations, cloud probability is defined as the number of time steps during which a parcel encounters saturation conditions as it moves forward in time along the path defined by the back-trajectory, divided by the number of time steps. These particular files show the probability over the preceding 24 hours.
d. Temperature and cloud probability: Same as (c) but with overlays of 375 hPa temperature (in red)
e. Cumulative cloud probability: Like (c) but over the entire time span of the trajectory
f. Temperature and cumulative cloud probability: Sames as (e) but with the temperarture overlay in (d)
g. Water vapor without dehydration: RDF water vapor fields unmodified by dehydration or resets to the local saturation mixing ratio in convectively-influenced regions along the trajectory path
h. Water vapor with dehydration: RDF water vaporfields with along-trajectory modification by dehydration or convective influence
i. Net water vapor change: The difference (h) minus (g)
j. Ozone at 375K: RDF fields as in (g)
k. Carbon monoxide at 375K: RDF fields as in (g)
For questions on the water vapor trajectory prouducts, contact Dr. Mark Schoeberl [email: mark.schoeberl@mac.com]