Elizabeth Frieberg
Meteorology 785
March 17, 2004
Question
A builder of custom homes in Southern California is in the process of developing plans for the construction of several high end houses in the San Gabriel Mountains. The sites of the homes are in hilly terrain and work on all of these properties will begin this June and take approximately five years to complete.
Although the drainage he will be
designing both for the homes and the property that they are on will be more than
adequate to handle even the 1000 year rainfall event, he is concerned about
construction delays that might occur due to rainfall over the next several
years. During construction, access roads will not be paved, and there will
be no vegetation to stabilize soils.
The builder has read that
scientists have determined that rainfall fluctuations over Southern California,
including the San Gabriel Mountains, seem statistically directly related to
the Pacific Decadal Oscillation (PDO), and that since PDO indices have been
negative and are expected to continue to be so for the next five years or so,
that sub-average precipitation can be expected during the same period. The
same report seems to downplay the relationship of El Niño to precipitation in
Southern California as dominant in the precipitation signal there.
Since this implies that the
builder will have to be less concerned about erosion on the side, or access
issues due to road washouts, he is interested in hiring a consulting
meteorologist:
1.
To determine if news reports about the PDO (and El Niño) and expected
rainfall in Southern California are
characterizing the findings of the science
with regard to these issues correctly; and
Introduction and Answer to Question 1
A major
aspect of the scientific method is reproducibility. A scientist must clearly outline his/her data used in the
study/case, along with the methodology that they followed in order to obtain
their results. When done properly,
another person should be able to come along and gather the same results as the
original scientist by following their procedure. In the attempts to reproduce the findings on the news report
for my client, major snags were hit. The main aspects that my client wished me
discuss are as follows with the conclusions that were obtained by examining the
actual data.
Top
PDO
“RAINFALL FLUCTUATIONS OVER SOUTHERN CALIFORNIA, INCLUDING THE SAN GABRIEL MOUNTAINS, SEEM STATISTICALLY RELATED TO PACIFIC DECADAL OSCILLATION (PDO)”
By using the same graph used in the news report and performing a statistical analysis, it is apparent that PDO is not statistically significant. The correlation coefficient is 0.25, which in order to be statistically significant, must be 0.05 or less. The significance explained is 0.06 is above this value, and it does not account for a good deal of the variance with only 6% explained, which makes PDO not a good indicator of rainfall in Southern California (including the San Gabriel Mountains).
Top
Seasonal Rainfall
“SINCE PDO INDICES HAVE BEEN NEGATIVE AND ARE EXPECTED TO BE SO FOR THE NEXT FIVE YEARS OR SO, SUB-AVERAGE PRECIPITATION CAN BE EXPECTED DURING THE SAME PERIOD.”
By
examining the graph of the 25-year running mean of PDO and Seasonal Rainfall,
notice that both the PDO and Seasonal rainfall are not in a negative phase, but
in fact have been steadily increasing. Even
if PDO were a statistically significant indicator of rainfall in Southern
California, this would indicate that both PDO and rainfall would be steadily
increasing.
Top
“REPORT DOWNPLAYS EL NINO AND SOUTHERN CALIFORNIA PRECIPITATION AS A DOMINANT FACTOR”
By performing the same statistical analysis on two other El Nino Indices (Southern Oscillation Index, and Nino3) as with PDO, we notice that both of these indices are statistically significant.
SOI
Top
Nino3
The
Correlation Coefficient for Nino3 is 0.55, which is a significance level of
0.01, which also makes it statistically significant.
Top
Conclusions and Answer to Question 2Both SOI and Nino3 are statistically significant, but they both only account for 27% of the variance. This is better than PDO (at 6%), but they are still not the best indicators of rainfall in Southern California.
In my report to the builder, I would convey my opinion that the news report that he heard were full of inconsistent information, and that PDO in fact is not the best indicator for Southern California precipitation. As to his construction timetable, I would suggest that he budget in time for expected rain delays due to mean seasonal rainfall for that region.
Depending on where he is planning on building the homes, he will have to account not only for seasonal rainfall, but also runoff from the higher peaks within the mountain range. Knowledge of the local watersheds is a good idea.
Top