l July 14, 2005

Research Images July 29, 2005


Preliminary tendency plots


Below is a recap of the timeseries of the (1) tropopause theta enkf analysis timeseries, (2) the tropopause theta field at 11/19 6 UTC (weakest point) and (3) 11/21 12 UTC (strongest point), (4) Timeseries of EnKF and GFS analyses for tropopause theta and (5) pressure:
Overview: Vortex at its weakest and strongest amplitudes



Soundings


Model soundings following pressure core
Core soundings
Temperature overlays of actual, WRF, and GFS profiles
YUX (Hall Beach, NT)
YRB (Resolute Bay, NT)
YCB (Cambridge Bay, NT)
Temperature difference profiles (T observed - T model). There were sometimes some pretty large differences. In tallying up some of the differences, the model seemed to have some of the biggest differences near the surface, and in general was too cold near the surface. The biggest differences were from the YRB (Resolute Bay) ones as I'm sure you'll notice (they're huge!). From 900 hPa - 500 hPa, of all the differences plotted below, I eyeballed that the model was too warm 20 times, too cold 7 times, and had neglible differences with the observed sounding 2 times. The same was true above 500 hPa, and I count (again just by eyeballing it) that overall the model was too warm 14 times, too cold 9 times, and it was hard to say 6 times.
YUX (Hall Beach, NT)
YRB (Resolute Bay, NT)
YCB (Cambridge Bay, NT)



Cross sections


Cross sections of relative humidity (colors), vertical motion (red = rising, blue = descending), and potential vorticity (black solid lines, contours = 0.5, 1, 1.25, 1.5, 2, 2.5, 3, 5). The one with the 'ew' in the name of the image is the east-west cross section and the one with the 'ns' in the name of the image is the north-south cross section:
Relative humidity, vertical motion (rising motion in red, descending blue), pv (contours at 0.5, 1, 1.25, 1.5, 2, 2.5, 3, 5)
Precipitation (reflectivity as in Ryan's Matlab script). Note the colorbar is NOT fixed:
Precipitation (Reflectivity) loop

Covariance Plot summary


Select vertical correlations between state variables and tropopause theta
State on 11/19/2004 6 UTC and metric on 11/19/2004 6 UTC
State on 11/19/2004 6 UTC and metric on 11/21/2004 12 UTC
Select sensitivities between state variables and tropopause theta. This is calculated by taking the covariance of the state variable and tropopause theta averaged within the 280K contour, dividing by the variance of the state variable, and multiplying by the ensemble standard deviation of the state variable inside the vortex core. The units are in Kelvin.
Below is for both the state variable and tropopause theta on 11/19/2004 at 6 UTC. This was when the vortex was weakest and was therefore about to strengthen rapidly. The first plot shows changes in tropopause theta for a 1 standard deviation change in temperatures at 800 hPa everywhere on the domain. The second plot is the corresponding east-west cross section. These plots show that in order to strengthen the vortex (decrease tropopause theta), temperatures below the tropopause must decrease and temperatures above must increase. The planar view shows the most sensitive region is at the vortex core and the cross section points out that the most sensitive levels are at 800 and ~300 hPa. Image 3 shows the changes in tropopause theta for a 1 standard deviation change in the v-component winds at 650 hPa. The corresponding cross section is Image 4. Both plots are relatively straightforward to interpret, as they say that increases in northerly winds everywhere will strengthen the vortex to the east side and weaken it to the west. Note the antisymmetry in the cross section (vortex trop. theta core is the green vertical line) is probably because the vortex pressure core is not collocated with the tropopause theta core. Images 5 and 6 show the changes in tropopause theta given a 1 standard deviation change in water vapor mixing ratios everywhere at 300 hPa and 800 hPa respectively. Increasing the water vapor above the tropopause can strengthen the vortex in places and in general strengthens it more than weakens it. However, increasing it below the tropopause mostly weakens the vortex at the center but strengthens it on the edges. Image 7 shows the planar view of changes in tropopopause theta with a 1 standard deviation change in relative humidity at 950 hPa. Image 8 shows the corresponding cross section. The interpretation is qualitatively the same as for water vapor mixing ratios, where low level moisture increases weaken the vortex and upper level increases can strengthen it. The most sensitive region is on the western edge of the vortex (where there is precipitation falling) and in the low levels. I think these would be consistent with our recent findings in that the closer to the tropopause (i.e. higher up) diabatic heating processes are occurring, the better the chances of its strengthening. This is more likely to occur when moisture gets wrapped closer into the center where the tropopause is lower.
State on 11/19/2004 6 UTC and metric on 11/19/2004 6 UTC
Below is for the state variable on 11/19/2004 at 6 UTC and tropopause theta on 11/21/2004 at 12 UTC. This was when the cyclone was weakest and strongest respectively. The first image shows the changes in tropopause theta on 11/21 at 12 UTC from changing the relative humidity by 1 standard deviation on 11/19 at 6 UTC everywhere. The most sensitive region closest to the 11/19 vortex core is to the east of Cambridge Bay. At this time, there is a maximum in water vapor mixing ratios centered at this same location. Image 2 shows the same as above except with changing the relative humidity by 1 standard deviation at 700 hPa. The results are different however, with the most sensitive region right in between the two vortex cores, over northwestern Baffin Island. On 11/19 at 6 UTC, there is a maximum in ice mixing ratios at that location (a loop of QICE is on the next link as a reference). I think all this says is that tropopause theta on 11/21 is most sensitive to the water vapor and ice mixing ratios where they are largest and closest to the core. Image 3 are changes in tropopause theta on 11/21 12 UTC from 1 standard deviation changes in temperature on 11/19 6 UTC everywhere. Here, the most sensitive region near the vortex cores is in northern Baffin Island, where the QICE maximum is located. Image 4 is the same as above except for temperature at 400 hPa. The pattern reflects the same as above but also strongly (relatively speaking) reflects how increasing the temperatures in the stratosphere implies a lower tropopause.
State on 11/19/2004 6 UTC and metric on 11/21/2004 12 UTC
Below is the total vertically integrated ice mixing ratios from 11/18/2004 00 UTC - 11/23/2004 00 UTC. The loop for the cloud water mixing ratios shows a negligible field, indicating that any clouds are primarily in the form of ice, so phase changes would be primarily between vapor and ice.
Total integrated QICE loop