Modeling rainfall and flooding

Xnn

I will not comment about earthquakes and tsumani's.
However, there has been an increase in extreme precipitation
events due to global warming/greenhouse gases.

Basically, greenhouse gases have increased the ability
of the atmosphere to hold water vapor. It is not that
there are more storms than in the past, instead they are more intense.

http://www.ipcc.ch/pdf/assessment-re...1-chapter3.pdf

Tropical cyclones making landfall in China are
a small fraction of the total storms, and no obvious long-term
trend can be discerned (He et al., 2003; Liu and Chan, 2003;
Chan and Liu, 2004). However, Emanuel (2005a) and Webster
et al. (2005, 2006) indicated that the typhoons have become
more intense in this region, with almost a doubling of PDI values
since the 1950s and an increase of about 30% in the number of
category 4 and 5 storms from 1990 to 2004 compared with 1975
to 1989.

Substantial increases are found in heavy precipitation
events. It is likely that there have been increases in the number
of heavy precipitation events (e.g., 95th percentile) within many
land regions, even in those where there has been a reduction in
total precipitation amount, consistent with a warming climate
and observed significant increasing amounts of water vapour
in the atmosphere. Increases have also been reported for rarer
precipitation events (1 in 50 year return period), but only a few
regions have sufficient data to assess such trends reliably.

Precipitation has generally increased over land north of
30°N over the period 1900 to 2005 but downward trends
dominate the tropics since the 1970s. From 10°N to 30°N,
precipitation increased markedly from 1900 to the 1950s, but
declined after about 1970. Downward trends are present in the
deep tropics from 10°N to 10°S, especially after 1976/1977.
Tropical values dominate the global mean. It has become
significantly wetter in eastern parts of North and South America,
northern Europe, and northern and central Asia, but drier in the
Sahel, the Mediterranean, southern Africa and parts of southern
Asia. Patterns of precipitation change are more spatially
and seasonally variable than temperature change, but where
signifi cant precipitation changes do occur they are consistent
with measured changes in stream flow.

Droughts have become more common, especially in the
tropics and subtropics, since the 1970s. Observed marked
increases in drought in the past three decades arise from more
intense and longer droughts over wider areas, as a critical
threshold for delineating drought is exceeded over increasingly
widespread areas. Decreased land precipitation and increased
temperatures that enhance evapotranspiration and drying
are important factors that have contributed to more regions
experiencing droughts, as measured by the Palmer Drought
Severity Index. The regions where droughts have occurred
seem to be determined largely by changes in SSTs, especially
in the tropics, through associated changes in the atmospheric
circulation and precipitation. In the western USA, diminishing
snow pack and subsequent reductions in soil moisture also
appear to be factors. In Australia and Europe, direct links to
global warming have been inferred through the extreme nature
of high temperatures and heat waves accompanying recent
droughts.

Tropospheric water vapour is increasing. Surface specific
humidity has generally increased after 1976 in close association
with higher temperatures over both land and ocean. Total
column water vapour has increased over the global oceans by
1.2 ± 0.3% per decade from 1988 to 2004, consistent in pattern
and amount with changes in SST and a fairly constant relative
humidity. Strong correlations with SST suggest that total
column water vapour has increased by 4% since 1970. Similar
upward trends in upper-tropospheric specific humidity, which
considerably enhance the greenhouse effect, have also been
detected from 1982 to 2004.

mheslep

Quote by Xnn

I will not comment about earthquakes and tsumani's.
However, there has been an increase in extreme precipitation
events due to global warming/greenhouse gases.

Basically, greenhouse gases have increased the ability
of the atmosphere to hold water vapor. It is not that
there are more storms than in the past, instead they are more intense.

http://www.ipcc.ch/pdf/assessment-re...1-chapter3.pdf

That may be, but I don't believe the language you used (highlighted) is supported by the IPCC. They make a great many observations, and express concern:

Quote by IPCC AR4 3.8.1

There is increasing concern that extreme events may be changing in frequency and intensity as a result of human influences on climate

but do not state the causation as a fact.

Xnn

Theory and modeling both predict that hurricane intensity should increase with
increasing global temperatures and we already know that rising levels of CO2
cause rising global temperatures. However, the degree to which the rise
in extreme precipitation events and storms are due to the rise in greenhouse
gases has not been quantified. There are after all, natural variations some
of which span decades.

Here is a letter in the Journal Nature defining an index for potential destructiveness
of hurricanes that factors in both duration and intensity. There has been a marked
increase since the 1970's.

http://www.nature.com/nature/journal...ture03906.html

Theory and modelling predict that hurricane intensity should increase with increasing global mean temperatures, but work on the detection of trends in hurricane activity has focused mostly on their frequency and shows no trend. Here I define an index of the potential destructiveness of hurricanes based on the total dissipation of power, integrated over the lifetime of the cyclone, and show that this index has increased markedly since the mid-1970s. This trend is due to both longer storm lifetimes and greater storm intensities. I find that the record of net hurricane power dissipation is highly correlated with tropical sea surface temperature, reflecting well-documented climate signals, including multi-decadal oscillations in the North Atlantic and North Pacific, and global warming. My results suggest that future warming may lead to an upward trend in tropical cyclone destructive potential, and—taking into account an increasing coastal population—a substantial increase in hurricane-related losses in the twenty-first century.

Xnn

Here's another paper that has found an increase in floods during the 20th century
due to global warming with the expectation that the trend will continue.

http://www.nature.com/nature/journal...l/415514a.html

Radiative effects of anthropogenic changes in atmospheric composition are expected to cause climate changes, in particular an intensification of the global water cycle with a consequent increase in flood risk. But the detection of anthropogenically forced changes in flooding is difficult because of the substantial natural variability; the dependence of streamflow trends on flow regime, further complicates the issue. Here we investigate the changes in risk of great floods—that is, floods with discharges exceeding 100-year levels from basins larger than 200,000 km2—using both streamflow measurements and numerical simulations of the anthropogenic climate change associated with greenhouse gases and direct radiative effects of sulphate aerosols. We find that the frequency of great floods increased substantially during the twentieth century. The recent emergence of a statistically significant positive trend in risk of great floods is consistent with results from the climate model, and the model suggests that the trend will continue.

Evo

Quote by Xnn

Here's another paper that has found an increase in floods during the 20th century
due to global warming with the expectation that the trend will continue.

http://www.nature.com/nature/journal...l/415514a.html

What?

Please point out the floods it has cited, the specific number of floods, location, and how they positively linked this to "global warming" by ruling out any possibility of natural occurance. Such as natural or man made changes to the terrian that could cause flooding. Why don't I find anything about an increase of actual floods?

This appears to be nothing more than maybe, might, possibly, perhaps...

Where are you reading this stuff you stated as fact?

mheslep

Thanks for the pointer to Emanuel; I'd heard of it like everyone else but never reviewed it.

Quote by Xnn

Theory and modeling both predict that hurricane intensity should increase with increasing global temperatures and we already know that rising levels of CO2 cause rising global temperatures. However, the degree to which the rise in extreme precipitation events and storms are due to the rise in greenhouse gases has not been quantified.

I don't have access to the full text at the moment, perhaps Emanuel does quantify? But without some kind of quantification of the effect, of what use is the prediction? Suppose (and I have no idea) the effect is predicted to be a cyclone wind speed increase of 1 part in 10000 for a 10 deg K rise in SST?

Quote by Xnn

There are after all, natural variations some
of which span decades.

Here is a letter in the Journal Nature defining an index for potential destructiveness
of hurricanes that factors in both duration and intensity. There has been a marked
increase since the 1970's.

http://www.nature.com/nature/journal...ture03906.html

Here are some comments on Emanuel 2005 from W. Grey, submitted to Nature in response:

Quote by Gray

The near universal references to the above paper by most of the major US media outlets and blogs since Katrina and Rita made US landfall requires a response from a few of us who study hurricanes. Having been involved with hurricane research and forecasting for nearly 50 years, I feel I have an obligation to offer comments on this paper's findings which, in my view, are not valid. This paper concludes that global tropical cyclone net power dissipation (or friction times wind) taken to be proportional to the sum of each cyclone's individual 6-hour track period maximum wind speed (Vmax^3) has undergone large (more than doubled) increases over the last 30 years. The author associates these frictional energy dissipation increases with rising mean sea surface temperatures (SSTs) and implies that these SST increases may, in part, be related to human activity. If true, this is a very important finding that has great relevance as regards to the globe's future climate and future hurricane destruction. But, the author's apparent "blockbuster" results and his interpretation of his calculations are not realistic.

http://arxiv.org/ftp/physics/papers/0601/0601050.pdf

If you are interested, Emanuel informally answers some of his critics on his MIT web site here.
Skip down to:
5. Empirical Evidence for Increasing Tropical Cyclone Activity (and a response to its critics)
http://wind.mit.edu/~emanuel/anthro2.htm

Xnn

Quote by Evo

What?
Why don't I find anything about an increase of actual floods?

This appears to be nothing more than maybe, might, possibly, perhaps...

Where are you reading this stuff you stated as fact?

Notice the last sentance of the abstract reads as follows:

The recent emergence of a statistically significant positive trend in risk of great floods is consistent with results from the climate model, and the model suggests that the trend will continue.

Seems clear enough and no maybe about it.

Don't have the full paper with all the details (it's available, but not free,).
Have emailed the author for more infomation.

Xnn

Here's a link to a free version of the Nature article on increased risk of great floods:

http://www.gfdl.noaa.gov/bibliograph...es/pcm0201.pdf

Notice it is a 2002 paper.

Wonder if all the recent flooding has made the correlation stronger.

mheslep

Quote by Xnn

Here's a link to a free version of the Nature article on increased risk of great floods:

http://www.gfdl.noaa.gov/bibliograph...es/pcm0201.pdf

Notice it is a 2002 paper.

Wonder if all the recent flooding has made the correlation stronger.

What do you mean by "all the recent flooding"? Are there data or news reports showing global flooding since 2002 are above normal?

Evo

Quote by Xnn

Notice the last sentance of the abstract reads as follows:

The recent emergence of a statistically significant positive trend in risk of great floods is consistent with results from the climate model, and the model suggests that the trend will continue.

Seems clear enough and no maybe about it.

That's not saying that there "has been an increase in floods", it says "risk". As in "it hasn't actually happened", but there is a chance. That's not the same thing.

This is just a prediction, not fact.

mheslep

I've been quickly through the Milly 2002 paper since Xnn's post. Part of the paper concerns model predictions for the future, which is off topic for this thread, and part of the paper concerns observations of existing food data. I'm having some trouble with the latter.

So far I have this:
Milly et al and others in their references admit that that there's no detectable signal in small floods, so they look at major 100-year flood events, that is events that have a probability of 1% on a given year. They globally look at 29 large watershed sites spanning the 135 yr period 1865 to 1999. That would give them a maximum of 3915 site-years, but the data does not cover the full time period for all sites and they end up with 2066 site-years. From that data, if flooding were normal, we would expect to observer 20.6 100-year floods. They look at all this data, and find there were exactly 21 100-year events globally. I don't follow how that finding produces the abstract statement "We find that the frequency of great floods increased substantially during the twentieth century". Anybody? I'm still re-reading the 2nd half of Milly.

sylas

Quote by mheslep

I've been quickly through the Milly 2002 paper since Xnn's post. Part of the paper concerns model predictions for the future, which is off topic for this thread, and part of the paper concerns observations of existing food data. I'm having some trouble with the latter.

So far I have this:
Milly et al and others in their references admit that that there's no detectable signal in small floods, so they look at major 100-year flood events, that is events that have a probability of 1% on a given year. They look at 29 large global watershed sites spanning the 135 yr period 1865 to 1999. That would give them a maximum of 3915 site-years, but the data does not cover the full time period for all sites and they end up with 2066 site-years. From that data, if flooding were normal, we would expect to observer 20.6 100-year floods. They look at all this data, and find there were exactly 21 100-year events globally. I don't follow how that finding produces the abstract statement "We find that the frequency of great floods increased substantially during the twentieth century". Anybody? I'm still re-reading the 2nd half of Milly.

It is not the existence of 21 events that is the basis of the quoted sentence, but the fact that those 21 events occur disproportionally in the second half of the record.

Extract:

Under the assumption that flood events were independent outcomes of a stationary process, we used binomial probability theory to determine a probability of 1.3% of having 16 or more of 21 events during the second part of the record. For observations from an extratropical subset of the basins (see below), the corresponding probability is 3.5%, for 7 out of 8 flood events in the second half of the record. Supplementary analyses for shorter return periods (2–50 yr) did not reveal significant trends, but 200-yr flood frequency increased significantly.

The hypothesis in the paper is explicitly described in the final paragraph as "tentative", but the statistical property they measure seems clear enough. The open issues identified in the paper are described in the final paragraph as follows:

Our detection of an increase in great-flood frequency and its attribution to radiatively induced climate change are tentative. The frequency of floods having return periods shorter than 100 yr did not increase significantly. Potentially significant effects of measurement non-stationarity are not easily assessed. The forced signal and unforced variability in the model contain errors of unknown magnitude. Absent from the model are forcings such as solar variability, volcanic activity, land-cover change⁹, and water-resource development¹⁰, and potential biospheric feedbacks such as CO₂-induced stomatal closure¹¹ and water-stress-induced root extension¹². Especially evident from our study are the needs for improvements in simulation of tropical hydroclimate and continued commitment to stream-gauging programmes worldwide.

(added in edit) This paper is not only about making predictions. It is also about a measured increased in "great" floods, and that is the portion most relevant to the thread. The key finding is: We find that the frequency of great floods increased substantially during the twentieth century. This is not an increase since 2001 of course, but it does have a clear connection to a topic of increasing disasters in present times. The development of discussion into a larger view of "present times" seems a pretty normal kind of thread development; but I would not object to locking the thread if it is to be limited so tightly as to preclude even this kind of associated discussion. I agree that the OP was not well founded.

D H

Quote by Evo

That's not saying that there "has been an increase in floods", it says "risk". As in "it hasn't actually happened", but there is a chance. That's not the same thing.

This is just a prediction, not fact.

From the paper (emphasis mine):

Our detection of an increase in great-flood frequency and its attribution to radiatively induced climate change are tentative. The frequency of floods having return periods shorter than 100 yr did not increase significantly. Potentially significant effects of measurement non-stationarity are not easily assessed. The forced signal and unforced variability in the model contain errors of unknown magnitude. Absent from the model are forcings such as solar variability, volcanic activity, land-cover change, and water-resource development, and potential biospheric feedbacks such as CO2-induced stomatal closure and water-stress-induced root extension.

How did this paper get published in Nature?

Evo

Quote by D H

From the paper (emphasis mine):

Our detection of an increase in great-flood frequency and its attribution to radiatively induced climate change are tentative. The frequency of floods having return periods shorter than 100 yr did not increase significantly. Potentially significant effects of measurement non-stationarity are not easily assessed. The forced signal and unforced variability in the model contain errors of unknown magnitude. Absent from the model are forcings such as solar variability, volcanic activity, land-cover change, and water-resource development, and potential biospheric feedbacks such as CO2-induced stomatal closure and water-stress-induced root extension.

How did this paper get published in Nature?

I was really surprised at this also.

mheslep

Quote by sylas

It is not the existence of 21 events that is the basis of the quoted sentence, but the fact that those 21 events occur disproportionally in the second half of the record.

Yes thanks, on re-read I see that point now.

Extract:

Under the assumption that flood events were independent outcomes of a stationary process, we used binomial probability theory to determine a probability of 1.3% of having 16 or more of 21 events during the second part of the record. For observations from an extratropical subset of the basins (see below), the corresponding probability is 3.5%, for 7 out of 8 flood events in the second half of the record. Supplementary analyses for shorter return periods (2–50 yr) did not reveal significant trends, but 200-yr flood frequency increased significantly.

The hypothesis in the paper is explicitly described in the final paragraph as "tentative", but the statistical property they measure seems clear enough. The open issues identified in the paper are described in the final paragraph as follows:

Our detection of an increase in great-flood frequency and its attribution to radiatively induced climate change are tentative. The frequency of floods having return periods shorter than 100 yr did not increase significantly. Potentially significant effects of measurement non-stationarity are not easily assessed. The forced signal and unforced variability in the model contain errors of unknown magnitude. Absent from the model are forcings such as solar variability, volcanic activity, land-cover change⁹, and water-resource development¹⁰, and potential biospheric feedbacks such as CO₂-induced stomatal closure¹¹ and water-stress-induced root extension¹². Especially evident from our study are the needs for improvements in simulation of tropical hydroclimate and continued commitment to stream-gauging programmes worldwide.

(added in edit) This paper is not only about making predictions. It is also about a measured increased in "great" floods, and that is the portion most relevant to the thread.

Edit: Yes, though I question whether that changes the answer to the query: What is the statistical significance of the additional floods in the latter part of the date? That is, given a coin that I test for randomness, if I toss the coin 20 times and came up with 10 heads, and also found that I observed only 4 heads in the first 10 tosses and 6 heads came in the last 10 tosses, am I justified in making in conclusions about the randomness of the coin somehow changing during the second ten tosses?

Edit: I see the paper comment's on 'stationarity'; I'm rusty on stationary processes but I believe that pertains to my question above.

The key finding is: We find that the frequency of great floods increased substantially during the twentieth century.

Its the statement published in the abstract, I'm not sure it should be called the 'key' finding, given the conclusion.

mheslep

Quote by D H

From the paper (emphasis mine):

Our detection of an increase in great-flood frequency and its attribution to radiatively induced climate change are tentative. The frequency of floods having return periods shorter than 100 yr did not increase significantly. Potentially significant effects of measurement non-stationarity are not easily assessed. The forced signal and unforced variability in the model contain errors of unknown magnitude. Absent from the model are forcings such as solar variability, volcanic activity, land-cover change, and water-resource development, and potential biospheric feedbacks such as CO2-induced stomatal closure and water-stress-induced root extension.

How did this paper get published in Nature?

I don't see any problem with publishing a narrow observation of one aspect a complex system, given that the authors freely admit that there may be a number of various drivers for observations as they do in the conclusion you posted above.

HOWEVER, in my view the stand-alone and qualitative statement in the abstract, "We find that the frequency of great floods increased substantially during the twentieth century", should have had more context or should have been dropped. For instance, as I read the paper, it would be equally appropriate make the following statement in the abstract: "We find that the frequency of 100 year floods was as expected over the observed 139 year time period (X% confidence), with an increasing frequency trend during the twentieth century". Mentioning the 100 year flood level and the time period immediately draw the readers attention to the tentativeness of the data set.

sylas

Quote by mheslep

Yes thanks, on re-read I see that point now.

Edit: Yes, though I question whether that changes the answer to the query: What is the statistical significance of the additional floods in the latter part of the date? That is, given a coin that I test for randomness, if I toss the coin 20 times and came up with 10 heads, and also found that I observed only 4 heads in the first 10 tosses and 6 heads came in the last 10 tosses, am I justified in making in conclusions about the randomness of the coin somehow changing during the second ten tosses?

Edit: I see the paper comment's on 'stationarity'; I'm rusty on stationary processes but I believe that pertains to my question above.

I presume we are still okay discussing this? I don't mind if a new thread is made; although it seems to me that this paper is a pretty close development from the OP; if the thread originator would like to stick to post 2001 developments then by all means we can split the thread.

A "stationary" process is stochastic process without a trend; or more formally the probabilities remain the same over time (this is even stronger). What we have here is basically testing the null hypothesis; under the assumption of a stationary process they see how likely the observed distribution might be. If is it very unlikely, then that can be taken as evidence that the null hypothesis is falsified, and that the process of producing floods is not stationary, but has a real underlying trend.

There's a fair bit more involved, and some statistics to deal with independence assumptions that I haven't tried to follow in detail, plus also the comparison of the observations with the behaviour of climate models, which would suggest that the trend seen in observations should be expected to persist; and that is where you get a hypothesis for the scientific explanation for the observed increase.

My reaction to this paper is that there's nothing especially dubious or unusual about publication of a study like this, that tests ideas without claiming to have a clear proof, and which indicates in the conclusion the areas of uncertainty, the limits of the approach and the useful directions for future work. It's appropriately tentative, while still have useful results.

Its the statement published in the abstract, I'm not sure it should be called the 'key' finding, given the conclusion.

Fair enough. One of the key points, then.

Cheers -- sylas

Similar Threads for: Modeling rainfall and flooding
Thread	Forum	Replies
Why does the temperature goes up following rainfall (then gradually cools down)?	General Physics	2
High rainfall event	Earth	3
Too Much Rain and Flooding!	General Discussion	27
Flooding in the west	General Discussion	30
New England Flooding	General Discussion	9

Xnn Blog Entries: 2	Here's a link to a free version of the Nature article on increased risk of great floods: http://www.gfdl.noaa.gov/bibliograph...es/pcm0201.pdf Notice it is a 2002 paper. Wonder if all the recent flooding has made the correlation stronger.