Home » Posts tagged 'Climate/Weather'

Tag Archives: Climate/Weather

Cold Winters and Phony Baloney (at least in Missouri)

This week I returned to St. Louis after being out of town for some time. I was greeted by a chorus of moans and groans about the horrible winter. Such kvetching! Of course, it is easy for somebody who has been in warmer climes to pooh-pooh the harshness of the winter back home. So, I decided to look and see what the statistics say, and since that is the focus of this blog, to do a post on what I found. I’m going to look at the winter in St. Louis and in Kansas City. For weather statistics, winter begins December 1st and ends February 28th (or 29th in leap years). I’m writing on February 21, so the data for this winter extends only through 2/20/2019. One final note: for grammatical reasons, in what follows, “normal” means historical average (mean).

The weather service office in each location keeps its data in slightly different formats, so I will do one, then the other.

Winter 2018-2019 in St. Louis

First, let’s ask if it has been excessively cold in St. Louis this winter. According to the National Weather Service, the record low temperature in St. Louis is -22°F, which occurred 1/5/1884. The observed low this winter was -6°F, on 1/20/19: cold, but nowhere near the record. For the 82 days from 12/1/18 through 2/20/19, on 57 of the days the record cold for that date has been -5°F or colder. This year, the low temperature has been nothing like that.

Figure 1. Data source: NOAA, National Weather Service, St. Louis Forecast Office.

Well, you may say, perhaps the low temperature has not set records, but on most days it has been lower than normal. Figure 1 shows the daily observed low temperature compared to the normal low temperature for that date. The blue line shows the observed temperature for 2018-19, and the red line shows the normal low temperature for that date. The chart suggests that for much of the winter, the low temperature in St. Louis has actually been above normal. There have been a few cold outbreaks, but not record cold. The observed low temperatures over the period this winter have averaged 27°F. The normal low temperatures over the period have averaged 26°F. So guess what? The average low temperature this year has been about a degree above normal.






Figure 2. Data source: NOAA, National Weather Service, St. Louis Forecast Office.

Well, you may say, perhaps the low temperature has not been excessively low, but the daily high temperature has been colder than usual. It’s not the deep lows of the night that has gotten to us, it’s the fact that it hasn’t warmed during the day. Figure 2 shows the daily observed high temperatures for 2018-19 (blue line), and the normal high temperature for those dates (red line). The chart shows that during the cold outbreaks noted above, the high temperature has, indeed, been cooler than normal. But much of the winter has also had highs above normal. Over the period, the observed highs this winter have averaged 43°F, while over the period, normal highs averaged 42°F.

Winter 2018-19 in Kansas City

The National Weather Service Office in Kansas City does not seem to publish a data series that contains information similar to the one published by the office in St. Louis. I have used, instead, data from the Climate-at-a-Glance data portal. This data does not include daily values, only monthly averages. Plus, it only extends through the end of January. January 19 was the coldest day of this winter, however, so it is included. Data collection began in 1972-73.

Figure 3. Source: Climate-at-a-Glance.

Figure 3 shows the data, with the blue line representing the observed values, and the gray line representing the average. The average temperature in Kansas City this winter was 2.5°F above normal.

The month of February to date can be included by using heating degree days instead of temperatures. Heating degree days are a measure designed to indicate to what degree the interior of buildings will require heating. To calculate it, average a day’s high and low temperature, then subtract the result from 65. This is how many heating degree days there were on that day. Now, to measure a period of time, simply sum the heating degree days for each day in the period.

The problem here is that the data in the climate summaries, where the heating degree data is published, use a different period to determine normal than does the data above. The data above uses values that run from when record keeping started to the current date. The climate summaries use data from 1981-2010. It was around 1980 that the effects of climate change really kicked in. This results in different estimates of “normal,” with the climate summary referencing only recent (warmer) history, and the other data referencing much longer (cooler) periods of time.

That said, it is the only way I can think of to include February for Kansas City in this discussion, so this is what the data shows:

Observed Heating Degree Days Normal Heating Degree Days Difference
December 2018 928 1040 -112
January 2019 1135 1114 21
February 1-20 2019 731 657


Looked at this way, it would appear that December created about 11% fewer degree days than normal, but January and February (to date) have created about 2% and 11% more, respectively. If you sum the differences for the 3 months together, then the winter to date has created 17 more heating degree days than normal, a trivial amount: in terms of heating degree days, Kansas City’s winter in 2018-19 should be understood to be roughly normal.

Now, none of this speaks to snow or blizzards. I understand that the winter storm at the end of January was a terrible event. In a similar fashion, I was in Hawaii when the winter cyclone came ashore in early February. I saw whole fields of banana trees leveled, just snapped off mid-trunk. On the top of Mauna Kea, the wind was recorded at 190+ mph. None of that changes the fact, however, that Hawaii has a lovely climate, and it was a wonderful place to visit (although too crowded these days, I’d say). Same in St. Louis. This blog is more concerned with statistical trends than individual events, and none of the statistics suggest that this has been, on average, a freakishly cold winter.

I read that people who believe in climate change are being peppered with the question “If the Earth is warming so much, how come it is so cold?” Nobody ever said that climate change would banish all cold, and the predictions are for more intense storms, just like the ones referenced above. But the real answer seems to be that it isn’t actually so cold, at least not here in Missouri. The whole question is nothing but phony baloney, at least here in Missouri.


NOAA National Centers for Environmental information, Climate at a Glance: U.S. Time Series, retrieved on February 21, 2018 from http://www.ncdc.noaa.gov/cag.

NOAA, National Weather Service, Kansas City/Pleasant Hill Forecast Office. 2/21/2019. Daily Climate Report. For this post, I used reports for 12/31/2018, 1/31/2019, and 2/20/2019. Viewed online 2/21/2019 https://w2.weather.gov/climate/index.php?wfo=eax.

NOAA, National Weather Service, St. Louis Forecast Office. 2/21/2019. Climate Graphs. Data retrieved on 2/21/2019 from https://www.weather.gov/lsx/cliplot.

Very Dry vs. Very Wet Months in the United States, 2018 Update

I’ve reported on drought in the American West many times in this blog. What about the country as a whole?

One way of looking at this question is by asking each month how much of the country has been very dry, and how much as been very wet? By very dry, I mean that the amount of precipitation for that month falls in the lowest 10% for that month in the historical record. By very wet, I mean that the amount of precipitation for that month falls in the highest 10% for that month.

The National Oceanic and Atmospheric Administration keeps this data. They measure the precipitation in every county in the country, and calculate what percent of the country was very dry, and what percent was very wet. They have data for every month going back to January of 1895.

Figure 1. Data source: National Centers for Environmental Information.

Figure 1 shows the monthly data for every month all the way back to January, 1895. Blue bars represent the percentage of the country that is very wet. Red bars represent the percentage that is very dry. (To keep the blue and red bars from obscuring each other, I multiplied the dry percentage by -1, thereby inverting it on the chart.) I dropped trend lines on both data series. As you can see, there is considerable variation from year-to-year. There is a slight trend – hardly noticeable – towards more very wet months and fewer very dry months. But it is small, and the yearly variation is much greater than the trend.



Figure 2. Data source: National Centers for Environmental Information.

Figure 2 shows the same data, but it beings in January, 1994.. I constructed this chart to see whether the most recent 25 years look different than the record as a whole. Again, blue bars represent very wet months, and the red bars represent very dry ones. I dropped linear trend lines on both data series, as before. The yearly variation is again larger than the trends. There appears to be virtually no trend in the number of very dry months. There is a small trend towards increasing number of very wet months. It appears a bit larger than did the one for the whole time period, but even so, it is tiny compared to the yearly variation.



Figure 3. Data source: National Centers for Environmental Information.

It’s a bit hard to read the two data series on opposite sides of the zero line, so I constructed Figure 3. For each month it shows the percentage of the country that was very dry minus the percentage that was very wet. By doing my subtraction that way, numbers above zero mean that more of the country was very dry than very wet, and numbers below zero mean that more of the country was very wet. I dropped a linear trend on the data (red), and I also dropped a 15-year moving average on it. The chart shows that, as we saw in Figure 1, there is a slight trend towards fewer very dry months and more very wet ones. The variation is much larger than the trend, whether one looks at the monthly data, or the yearly.

This data differs from other drought data I report. Those reports focus on the Palmer Drought Severity Index, an index intended to represent soil moisture. Soil can dry out because there is little overall precipitation, or because there are longer periods between precipitation events, or because the temperature is warmer. This data would tend to indicate that regions of the country with very little precipitation may be decreasing very slightly, very slowly. Regions with very much precipitation may be increasing. This trend would be consistent with consensus predictions regarding climate change, where overall precipitation is not expected to change, but the number of heavy precipitation events is expected to increase.


National Centers for Environmental Information, National Oceanographic and Atmospheric Administration. U.S. Percentage Areas (Very Warm/Cold, Very Wet/Dry). Downloaded 9/1/2018 from https://www.ncdc.noaa.gov/temp-and-precip/uspa.

Drought in American West and Southwest

Drought has one again gripped the American West and Southwest. Unfortunately, the wet winter of 2017 turned out to be a one-year reprieve. Perhaps the coming winter will be wet again, but for now, the regions are once again dry – in some cases, as dry as they have ever been since record-keeping began.

Figure 1. Source: Riganti, 2018.

Figure 1 shows the U.S. Drought Monitor for July 17, 2018. This map shows the Palmer Drought Severity Index for the United States. White areas are not in drought, colored areas are, and the darker the color, the worse the drought. It is easy to see that an “exceptional drought” has gripped portions of the Southwest, centered on the Four Corners Area where Colorado, New Mexico, Arizona, and Utah meet. Though the drought is most severe there, it has gripped much of the entire western United States.

Drought is primarily about soil moisture. Without moisture in the soil, crops wither and drinking water sources dry up. It is not feasible, however, to directly measure soil moisture across the entire country. The Palmer Drought Severity Index computes an estimate of soil moisture using temperature and precipitation data, and is the primary measure of drought used by the National Oceanographic and Atmospheric Administration.


Figure 2. Source: National Oceanographic and Atmospheric Administration, 2018.

Figure 2 shows the PDSI in June for the Southwest Climate Region (Arizona, Colorado, New Mexico, and Utah) from 1895 to 2018. Green columns represent years when it was wetter than average in June, orange columns when it was dryer than average. It is easy to see that green columns cluster to the left of the chart, and orange ones cluster to the right. In fact, of the most recent 19 years, 16 have been drier than average. This year, June was the second driest in the record, virtually as dry as it has ever been since record keeping began. The blue line shows the trend: the PDSI has decreased -0.23 per decade on average.

Ever since I wrote an extended series of posts on drought in California in 2015, this blog has followed the drought situation there. Plentiful snow during the winter of 2017 brought welcome relief, but the winter of 2018 was a big disappointment. Precipitation was below average, and the snowpack peaked well below normal (see here).

Figure 3. Source: Riganti, 2018.

As Figure 3 shows, California has not escaped the drought gripping most of the West. The most extreme drought is in the southeastern corner of the state. That is where the Imperial Valley is located, a region that supplies us with many of our fruits and vegetables. The farms there are mostly irrigated with water from the Colorado River, so local drought there is not a terrible worry for us here in Missouri. More on the Colorado River below, however.





Figure 4. Source: National Oceanographic and Atmospheric Administration, 2018.

Figure 4 shows that the PDSI for the state as a whole for June 2018, indicates a severe drought, but not an extreme one. The trend over time is clearly downward, however, and the continued dryness represents a long-term threat to the state’s water supply.






Figure 5. Source: California Department of Water Resources, 2018.

As Figure 5 shows, California’s reservoirs are moving into deficit. They have been fuller than average for most of the time since the winter of 2017, but now three of the biggest and most important, Trinity Lake, Lake Shasta, and Lake Oroville, are below average for this date. In the chart, the yellow bars represent the maximum capacity of each reservoir, the blue bars the current level, and the red line the average for this date. Lake Oroville has not yet fully recovered from the near disaster in 2017 that caused managers to lower the lake level to prevent a collapse of the dam(see here).








Figure 6. Source: Lake Mead Water Database, 2018.

Readers who have been following this blog know that California, Arizona, Nevada, Utah, New Mexico, and even Mexico depend on water from the Colorado River, and that Lake Mead is the large reservoir that holds the water for all of those states. You also know that water withdrawals from Lake Mead have exceeded inputs for many years, and the level of water in the lake has been relentlessly dropping. One study went so far as to predict that Lake Mead had a 50% chance of going dry by 2021. (Barnett and Pierce, 2008. See here for a fuller discussion California’s water resources.) Figure 6 shows the level of the lake for the past 3 years. The green line shows 2016, the red line 2017, and the blue line the year-to-date in 2018. Notice that the chart begins October 1, which is the official start of the water year. The wet winter of 2017 replenished the lake a little bit, but you can see that the current drought is causing it to drop again. Lake Mead is only 38% full, and Lake Powell, the large reservoir upstream from Lake Mead, is only 51% full. These low levels do not represent an immediate existential threat, but if dry conditions persist, they will before too many years pass.


Figure 7. Source: National Oceanographic and Atmospheric Administration, 2018.

The situation is different for Missouri. The most important source of water in our state is the Missouri Rivers (see here). Going back to Figure 1 above, you can see that drought is not severely impacting most of the region drained by the Missouri. As Figure 7 shows, precipitation in the Northern Rockies and Plains Climate Region was slightly above average for the first 6 months of 2018. Statistics for the reservoirs along the Missouri show that they are at or near maximum storage. In fact, since part of their mission is flood control, several of them are higher than desired for that purpose. (U.S. Army Corps of Engineers, 2018)

The real concern is that the drought in the American West might not be not a temporary weather phenomenon, but, instead, a permanent change in climate. Modelers predict that climate change will cause just such a change, could it be occurring already?


Barnett, Tim, and David Pierce. 2008. “When Will Lake Mead Go Dry?” Water Resources Research, 44, W03201. Retrieved online at http://www.image.ucar.edu/idag/Papers/PapersIDAGsubtask2.4/Barnett1.pdf.

Lake Mead Water Database. 2018. Lake Mead Daily Water Levels, Last 3 Water Years. Downloaded 7/19/2018 from http://graphs.water-data.com/lakemead.

National Oceanographic and Atmospheric Administration. 2018. Climate-at-a-Glance. Data downloaded 2018-07-19 from https://www.ncdc.noaa.gov/cag.

Riganti, Curtis. 2018. U.S. Drought Monitor, July 17, 2018. National Drought Mitigation Center. Downloaded 7/19/2018 from http://droughtmonitor.unl.edu.

U.S. Army Corps of Engineers, Missouri River Basin Water management. 2018. Mainstem and Tributary Reservoir Bulletin, 7/18/2018. Viewed online 7/19/2018 at http://www.nwd-mr.usace.army.mil/rcc/reports/pdfs/MRBWM_Reservoir.pdf.

Below Average Snowpack in the American West

The western snowpack was seriously below average this year, and it was way below average in the Lower Colorado Region.

It is early April, and that means it is time to check-in with snowpack data in California and the American West. On average, the snowpack reaches its maximum by April 1, after which it begins to shrink as it melts away. California and much of the West have a monsoonal precipitation pattern: the bulk of the yearly precipitation falls during the winter. Because the summer and fall are so dry, many regions depend on melting snow, which they collect into reservoirs. The snowpack serves as a kind of natural reservoir, collecting precipitation during the winter, and releasing it gradually as the snow melts.

Snowpack is measured in inches of water equivalent. To equal an inch of melted water requires between 7 and 20 inches of snow, depending on how slushy or powdery the snow is. To quantify the snowpack, scientists calculate how many inches of snow are on the ground, and how much water it would represent if it were instantaneously melted. The result is called the snow water equivalent. Thus, 1 inch of snow water equivalent means that, no matter how deep the snow is lying on the ground, if you melted it, it would equal 1 inch of water.

Figure 1. Source: California Department of Water Resources, California Data Exchange Center.

Figure 1 shows the snowpack in California for the three major snow regions: North, Central, and South, with the snow water equivalent given along the vertical axis on the left. The dark blue line represents the 2017-2018 winter, and the line ends on March 29. The blue number at the end of each blue line represents the snow water equivalent of this year’s snowpack as a percentage of the historical average for that date. At lower right the three regions are combined into a single number, representing the snow water content of the entire state’s snowpack for 3/29/18. At the bottom left the chart shows the statewide percentage compared to what’s average for April 1.

Through the end of February, this winter was the second driest on record, and the snowpack was something like 20% of average. March was a wet month, however, tripling the snowpack. Even so, that only brought it up to a statewide average of 57%.

Figure 2. Source: National Resources Conservation Service.

California also depends on water from outside of the state, especially water from the Colorado River. Figure 2 shows readings for the entire region upon which California draws. It  encompasses much of the southwestern United States. The data for this map come from a different data set than the ones in the previous chart, and thus the data for California are slightly different. (Most of the difference probably arises from using somewhat different reference periods to represent “average.”)

As you can see, the entire region has had a smaller than average snowpack. However, the snowpack in the Lower Colorado Region is particularly worrisome, as it is only 21% of average.



Figure 3. Data source: Mammoth Mountain Ski Resort.

The Mammoth Mountain Ski Resort publishes a detailed history of the snowfall at the resort, and I use it as an example of the snowfall in a given California location. Figure 3 shows the data. The total amount of snow at Mammoth Mountain through March 31 was 248 inches this year, compared to an average of 308 over the period from 1969-2018. The length of the colored bars for 2018 illustrates that more than half of the snow for the whole season fell during March. The chart also shows just how wet a winter it was last year, the second wettest in the record. Bear in mind that Mammoth Mountain is measuring snowfall, not snowpack.

So, measurements of the snowpack indicate that it is seriously below average. What, then, is the status of California’s water supply? The quick answer is that for this year they should be fine.

California’s water supply is impacted this year by an extraordinary circumstance: in February, 2017, the Oroville Dam suffered a failure of the main and emergency spillways, leading to the evacuation of 188,000 people lest the dam fail entirely (see here). It didn’t fail, but since then the reservoir has been partially emptied to facilitate repairs and improvements.

Figure 4. Source: California Department of Water Resources, California Data Exchange Center (A).

Figure 4 shows the data for the largest California reservoirs. On the chart, the blue bars represent the level of each reservoir on March 30, while the yellow bars represent the maximum capacity. The red line represents the historical average level of each reservoir on March 30. The blue number below the bars represents the amount of water in each reservoir compared to its capacity, while the red number represents the amount of water compared to the historical average for March 30.

As you can see, most of the reservoirs are at or above their average for March 30, and only Lake Oroville is considerably below average. The region around Santa Barbara, however, remains in a serious drought. The two largest reservoirs in Santa Barbara County, the Cachuma and Twitchell Reservoirs, are at 40% and 2% of capacity, respectively (not shown on the chart).





Figure 5. Source: lakemead.water-data.com.

In addition to the California reservoir system, southern California relies heavily on water from the Colorado River. Lake Mead, the largest reservoir on the Colorado River, has been overused for years, and was even forecast to have a strong chance of going dry (see here). Figure 5 plots the water level at Lake Mead over the past year. Each year it fills with the spring snowmelt, and then is drawn down throughout the rest of the year. Beginning just after 2000 Lake Mead has suffered a steady and rather alarming drop. Last year, for the first time in many years, Lake Mead showed a year-to-year increase in its water storage. This year, as of April 1, the water level of Lake Mead is basically unchanged from last year.

Lake Powell, a large reservoir upstream from Lake Mead, is up 16 feet from last year on this date. That is a significant increase, and it comes entirely from the large snowpack last year.

So, what does all this mean? The snowpack this year was seriously below average, and it was way below average in the Lower Colorado drainage region. California’s reservoirs, however, appear to be in good shape except in the region around Santa Barbara. Lake Mead has not lost additional water, and the fact that Lake Powell has gained water means that officials may be able to move water from there to Lake Mead if needed. Thus, the water supply, for this year may be sufficient for California and for those regions that draw on the Colorado River below Lake Mead.

It is worrisome, however, that after having experienced a severe multi-year drought, and then only 1 year of high precipitation, California and the Southwest have returned to below average snowpacks. I have reported previously that climate predictions include a permanent reduction of the snowpack throughout the West (see here) and in California (see here). We will have to keep watching over many years to see how this plays out.


California Department of Water Resources, California Data Exchange Center. Reservoir Conditions, 4/1/2018. Downloaded 4/2/2018 from http://cdec.water.ca.gov/cgi-progs/products/rescond.pdf.

California Department of Water Resources,         California Data Exchange Center. California Statewide Water Conditions, Current Year Regional Snow Sensor Water Content Chart (PDF). Downloaded 4/1/2018 from https://cdec.water.ca.gov/water_cond.html.

lakemead.water-data.com. Lake Mead Daily Water Levels. Downloaded 4/1/2018 from graphs.water-data.com/lakemead.

Mammoth Mountain Ski Resort. Snow Conditions and Weather. Viewed online 4/1/2018 at https://www.mammothmountain.com/winter/mountain-information/mountain-information/snow-conditions-and-weather.

National Resources Conservation Service. Open the Interactive Map. Select “Basins Only.” On the map, select “Percent oNCRS 1981-2010 Average,” “Region,” “Watershed Labels,” and “Parameter.” Downloaded 4/2/2018 from https://www.wcc.nrcs.usda.gov/snow/snow_map.html.

Santa Barbara County Flood Control District. Rainfall and Reservoir Summary, 4/1/2018. Viewed online 4/2/2018 at https://www.countyofsb.org/uploadedFiles/pwd/Content/Water/Documents/rainfallreport.pdf.

Above Average Precipitation in 2017 for Contiguous USA

2017 was the 19th wettest year on record across the contiguous USA.

Figure 1. Source: NOAA, Climate-at-a-Glance.

So says data from Climate-At-A-Glance, the data portal operated by the National Oceanographic and Atmospheric Administration (NOAA). Figure 1 shows the data, with the green line representing actual yearly precipitation, and the blue line representing the trend across time. The left vertical scale shows inches of precipitation, while the right shows millimeters of precipitation. In 2017, the average precipitation across the contiguous USA was 32.21 inches, which was the 19th highest amount in the record. Over time, precipitation seems to be increasing at about 0.17 inches per decade. The trend towards more precipitation is present in the Eastern Climate Region (+0.25 inches per decade), the Southern Climate Region (+0.22 inches per decade), and the Central Climate Region (+0.22 inches per decade). It is almost absent in the Western Climate Region, however (+0.03 inches per decade). (Except where noted, data is from the Climate-at-a-Glance data portal.)

(Click on figure for larger view.)

Figure 2. Source: NOAA, Climate-at-a-Glance.

In Missouri, 2017 was the 51st wettest year on record, with 41.22 inches of precipitation. (Figure 2) This puts the year slightly above the long-term average. As expected, the variation from year-to-year is much larger than the change in precipitation over time, but since 1895 Missouri has trended towards about 0.24 inches more precipitation per decade.

The interesting thing about Missouri’s precipitation is that in each of the last 2 years, concentrated storm systems have moved across the state from southwest to northeast, roughly following the route of I-44. They have led to huge amounts of rain over periods of a couple of days, resulting in damaging flooding. (See here and here.) This pattern is the one predicted by climate change models – slightly increased precipitation occurring in heavy precipitation events, with longer, drier spells between. (Drier because increased temperatures will cause the soil to dry out more quickly.)

Table 3. Source: NOAA, Climate-at-a-Glance.

The Northern Rockies and Plains are where most of the water that flows into the Missouri River originates, and the Missouri River provides water to more Missourians than any other source. This region saw 21.17 inches of precipitation in 2017, some 0.28 inches below average. (Figure 3) As expected, the variation between years is much larger than the change over time, but here, too, precipitation has been increasing, though the change has only been +0.07 inches per decade.

What to watch for in Missouri, then, does not appear to be a decrease in average yearly precipitation, but two other issues. First, demand for water has been increasing. Will it grow to outstrip the supply? Second, climate change is causing precipitation that once fell as snow to fall as rain. This changes the timing of when the Missouri River receives the runoff. Will that affect the ability of the river to supply water to meet the demand for water? So far, these answers are not known. (For a more extended discussion, see here.)

Figure 4. Source: NOAA, Climate-at-a-Glance.

The water situation in California is more serious than it is in the Northern Rockies and Plains, Missouri, or contiguous USA. California has a monsoonal precipitation pattern, and it has regions that receive a great deal of precipitation, while other regions receive little, if any. Consequently, the state relies on snowfall during the winter, which runs off during the spring and early summer, and is collected into reservoirs. This water is then distributed around the state. Thus, the amount of water contained in the snowpack on April 1, which is when it historically started melting in earnest, has been seen to be crucial to California’s water status.

After a severe, multi-year drought, last year was a big water year in California. (Figure 4) They received huge amounts of snow during January and February. For instance, the Mammoth Mountain Ski Area received 408 inches of snow during the 2 months. (Mammoth Mountain 2018) Over the whole year California received 27.63 inches of precipitation. That is the 22nd highest amount in the record, and it is 5.24 inches more than average.

Figure 5. Source: California Data Exchange Center, Dept. of Water Resources.

Unfortunately, this winter is not being as kind to California as last year, at least not so far. December, 2017, was the 2nd driest December on record, with only 1989 being dryer. The snowpack measurements suggest that the state has only about 22% of the snowpack that is average for this time of year (Figure 5, data as of 1/22/2018, California Snowpack Survey 2018) This is echoed by data from the Mammoth Mountain Ski Area, which reports only 73 inches of snow to date, vs. 349.5 inches through the end of January last year. (As I write, there are a few days left in January, but it still looks like a very serious shortfall to me.)

The snowpack is also below average in the Colorado River Basin above Lake Powell, the other major source for California’s water. As of 1/28/2018, the snowpack is only 65% of the average for this date. (National Resource Conservation Service, 1/28/2018) Now, snow tends to fall during storms, and there is no predicting when the storms will come. February and March could still bring much-needed snow. But California just got out of a terrible multi-year drought, and it would be very disappointing if it went right back into another after only 1 year.

ADDENDUM: A few days after I wrote this article, the New York Times published one on the water crisis in Cape Town, South Africa. That city is only about 3 months from running completely out of water. This blog focuses on statistics and big pictures. If you want a perspective on what such a crisis might actually look like in an urban area, I recommend the Times article.


California Data Exchange Center, Department of Water Resources. Current Year Regional Snow Sensor Water Content Chart (PDF). Downloaded 1/22/2018 from https://cdec.water.ca.gov/water_cond.html.

Mammoth Mountain Ski Area. 2018. Snow Conditions and Weather: Snow History. Viewed online 1/15/2018 at NOAA National Centers for Environmental information, Climate at a Glance: U.S. Time Series, published January 2018, retrieved on January 15, 2018 from http://www.ncdc.noaa.gov/cag.

Natural Resource Conservation Service, U.S. Department of Agriculture. Upper Colorado River Basin SNOTEL Snowpack Update Report. Viewed online 1/28/2018 at https://wcc.sc.egov.usda.gov/reports/UpdateReport.html?textReport.

NOAA National Centers for Environmental information, Climate at a Glance: U.S. Time Series, published January 2018, retrieved on January 15, 2018 from http://www.ncdc.noaa.gov/cag.

The Second Warmest Year Ever

2017 was the 2nd warmest year on record globally, and the 3rd warmest for the contiguous USA.

Figure 1. Data source: NOAA, Climate-at-a-Glance.

Figure 1 shows the average annual temperature for the Earth from 1880-2017. The chart shows the temperature as an anomaly. That means that they calculated the mean annual temperature for the whole series, and then presented the data as a deviation from that mean. Degrees Celsius are on the left vertical axis, and degrees Fahrenheit are on the right. Because the earth contains very hot regions near the equator and very cold polar regions, the actual mean temperature has relatively little meaning, and Climate-at-a Glance does not include it in their chart. (Except where noted, all data is from NOAA, Climate at a Glance.) 2016 was the highest on record, but 2017 was second. The 4 highest readings have all occurred within the last 4 years. You can see that the Earth appears to have been in a cooling trend until around 1910, then a warming trend until mid-Century, then a cooling period until the late 1960s or early 1970s, and then a warming period since 1970. Over the whole series, the warming trend has been 0.07°C per decade, which equals 0.13°F per decade. Since 1970, however, the warming has accelerated to 0.18°C per decade (0.32°F).

(Click on chart for larger view.)

Figure 2. Data source: NOAA, Climate-at-a-Glance.

Figure 2 shows the average yearly temperature for the contiguous United States from 1895 to 2017. In this chart and those that follow, the vertical axes are reversed, with °F on the left vertical axis, and °C on the right. The purple line shows the data, and the blue line shows the trend. 2017 was the 3rd highest in the record at 54.58°F. The 4 highest readings have all come within the last 6 years. Over time, the average temperature has increased 0.15°F per decade. Since 1970, however, the rate has increased to 0.52°F per decade.


Figure 3. Data source: NOAA, Climate-at-a-Glance.

Figure 3 shows the average temperature across Missouri for 2017. Across the state, it was the 8th warmest year on record, with an average temperature of 57.1°F. In Missouri, the warming trend from 1930-1950 was more moderate than it was nationally, and the trend has been for a 0.1°F increase in temperature each decade. Since 1970, however, the increase has accelerated to 0.4°F per decade.




Figure 4. Data source: NOAA, Climate-at-a-Glance.

Because conditions in the Northern Rockies and Plains affect how much water flows into the Missouri River, which provides more of Missouri’s water supply than any other source, I have also tracked climate statistics for that region. Figure 4 shows the data. Last year was the 11th warmest in the record at 44.9°F. This region has been warming at a rate of 0.2°F per decade over the whole period, but since 1970, the rate has accelerated to 0.5°F per decade.




Figure 5. Data source: NOAA, Climate-at-a-Glance.

Because I have been concerned about the water supply in California, I also track the climate statistics for that state. Figure 5 shows the data. Last year was the third warmest year in the record, with an average temperature of 60.3°F. California has been warming at a rate of 0.2°F each decade. Since 1970 the rate of increase has accelerated to 0.5°F per decade.

In all 4 locations the average yearly temperature seems to have increased significantly for several decades, then paused during mid-Century, and then resumed climbing, but at an accelerated rate. There seems to be little doubt that across the country it is warmer than it was. In Missouri, the average yearly temperature has been increasing, but at a rate that is somewhat less than in the other locations I looked at.


NOAA National Centers for Environmental information, Climate at a Glance: U.S. Time Series, published January 2018, retrieved on January 15, 2018 from http://www.ncdc.noaa.gov/cag.

Did Missouri Have Record Cold?

We’ve had some cold weather in Missouri recently. St. Louis hit -6°F on New Years Day, while Kansas City hit -11°F. But these are not records. The record low on New Years day is -10°F in St. Louis, and -13°F in Kansas City.

Kansas City’s all-time record low is -23°F, which occurred in December 1989.

Figure 1. Data source: National Weather Service, St. Louis Forecast Offices, Personal communication from Spencer Mell.

Figure 1 shows a chart for each winter (December, January, and February). Blue columns are the number of days with a low temperature at or below 0°F in St. Louis, and they run from 1874 to 2016. Red columns are for Kansas City, and they run from 1888 to 2016. The dashed blue line represents the trend over time for St. Louis, the dashed red line for Kansas City. You can see that the number of days varies widely from year-to-year. Many years have 1 day, or even none. In St. Louis the maximum number of days was 18, and it occurred in the winter that began in December 1935. In Kansas City, the maximum number of days was 19, and it occurred twice: in 1935 and 1978.

The trend lines show that in Kansas City, the number of days has not been changing over time. In St. Louis, however, the number of days has decreased over time.

(Click on figure for larger view.)

Figure 2. Data source: National Weather Service, St. Louis Forecast Offices, Personal communication from Spencer Mell.

One can count the number of winters that had 0 days below 0°F, the number of winters that had 1 day, the number of winters that had 2 days, etc. You can then construct a frequency chart of how many years had each number of days. Figure 2 shows such a frequency chart for St. Louis and Kansas City. There have been 54 winters in St. Louis when there were no days with lows at or below 0°F, there have been 28 such winters in Kansas City, and no other number is represented in more years than that.

The number of extremely cold days varies widely from year-to-year, but in St. Louis the average number is 3, and in Kansas City it is 4. St. Louis has experienced 2 days below 0°F this winter, and Kansas City has experienced 4 (both as of 1/16). For comparison, St. Louis has had more than 2 days below 0°F some 51 times since 1874. Kansas City has had more than 4 days below 0°F some 31 times since 1888.

The severe cold began this year on the morning of New Years Day. What about last year? Was it a hot one, or not so hot? The next post will review average temperatures for all of 2017.


National Weather Service, Kansas City Forecast Office. 2018. WFO Monthly/Daily Climate Data. Data viewed online 1/15/2018 at http://w2.weather.gov/climate/getclimate.php?date=&wfo=eax&sid=MCI&pil=CF6&recent=yes&specdate=2017-12-31+11%3A11%3A11.

National Weather Service, St. Louis Forecast Office. 2018. Ranked Occurrences of Temperature <= 32 and 0 Degrees (1893-Present). Downloaded 1/15/2018 from http://www.weather.gove/lsx/cli_archive. (Actually contains data back to 1874).

Personal communication from Spencer Mell, Climate Focal Point, National Weather Service, Kansas City Forecast Office.

Hurricane Harvey Devastates Houston

Hurricane Harvey caused record flooding in Houston. Those poor people!

Figure 1. Flooding in Port Arthur, TX, from Hurricane Harvey. Source: South Carolina National Guard, 2017.

Most of you know about the terrible disaster that Hurrican Harvey caused in Houston, TX. The disaster will inevitably be compared to Hurricane Katrina and the flood that struck New Orleans. In both cases, a major city was flooded by a hurricane. Houston, however, is a metropolitan area with a population of about 6.3 million people, while New Orleans is a metropolitan area with a population of about 1.3 million. That means that Houston is almost 5 times as large.

New Orleans flooded so catastrophically because much of the city is below sea level. The levies broke, the ocean poured through, and the low areas filled up with water just like a bathtub would. Coastal Texas is a flat, low-lying area, some of which was swamp or marshland before being developed. It is not below sea level, however. Houston flooded because Hurricane Harvey dumped prodigious amounts of rain on the city – more than 4 feet of rain in some areas. The water couldn’t run off fast enough, and flooding occurred. The tragedy has been well covered by all of the national news sources, so I have contented myself with a single photograph of the flooding in Port Arthur, a small city about 100 miles northeast of Houston. (Figure 1) This blog focuses not on individual events, but on trends and on the big picture.

(Click on photo for larger view.)

Figure 2. Flooding in Houston from Tropical Storm Allison in 2001. Source: Harris County Flood Control District.

Houston has been hit repeatedly by tropical storms and hurricanes. From 1836 to 1936, the city suffered through 16 major floods, with the water level reaching as high as 40 feet in one of them. Since 1935, there have been 8 more. In 2001, Tropical Storm Allison dumped up to 35 inches of rain on Houston over 5 days, resulting in flooding that damaged over 73,000 homes and caused $5 billion in property damage (see Figure 2). In 2008, Hurricane Ike passed directly over the city, breaking out windows in downtown skyscrapers and wiping out electricity to some customers for over a month. Over the Memorial Day Holiday in 2015, rain of up to 11 inches over 24 hours drenched Houston, flooding thousands of homes. In April 2016 (last year), a trough of rain parked over the city, and over 24 hours, 17 inches of rain fell. They had to rescue 1,800 people from the floods, but even so 8 died and 1,144 homes were inundated.

But flooding is not limited to Houston. In April of this year, flooding in Missouri and Arkansas caused $1.7 billion in damages. In February, flooding in California caused $1.5 billion in damages, including Oroville Dam (see here). In October, 2016, Hurricane Matthew churned along the Atlantic Coast causing damage. In August, 2016, Louisiana received 20-30 inches of rain from a stationary storm, causing $10.3 billion in damages. And a December 2015 storm brought record flooding to Missouri and tornadoes to Texas, causing 50 deaths and $2.5 billion in damages. The list goes on and on.

Figure 3. Hurricane Irma Devastated the Island of St. Maarten. Photo: Gerb van Es, Dutch Department of Defense.

UPDATE: As of 9/8/2017, three more tropical storms have formed in the Atlantic Ocean: Hurricane Irma, a Catagory 5 hurricane (the largest category), passed over several Caribbean islands causing terrible damage (see Figure 3). As I write, it is bearing down on Florida. How bad will it be? We don’t know; it has diminished to a Category 4 hurricane, but it is wider than the Florida Peninsula is, and it is currently forecast to travel south to north right up the entire peninsula. Tropical Storm Jose is gaining strength in the mid-Atlantic, threatening many of the same islands that were just devastated by Irma, though it is forecast to turn north. And Hurricane Katia has formed just north of the Yucatan Peninsula, and is expected to come ashore north of Veracruz, Mexico.

What is happening? Has it always been this way, or is there more very damaging weather than there used to be? The next post will look at the national trend, and the post after that will look at the trend in Missouri.


Gerb van Es, Dutch Department of Defense. Aerial Photo Shows the Damage of hurrican Irma in Phillipsburg, on the Dutch portion of the Caribbean Island of Sint Maarten. Downloaded 9/8/2017 from https://www.caymancompass.com/2017/09/07/enormous-catastrophe-st-martin-reeling-from-hurricane-damage.

Harris County Flood Control District. Harris County’s Flooding History. Viewed online 8/30/2017 at https://www.hcfcd.org/flooding-floodplains/harris-countys-flooding-history.

Harris County Flood Control District. Tropical Storm Allison. Viewed online 8/30/2017 at https://www.hcfcd.org/storm-center/tropical-storm-allison-2001.

NOAA National Centers for Environmental Information (NCEI) U.S. Billion-Dollar Weather and Climate Disasters (2017). https://www.ncdc.noaa.gov/billions.

South Carolina National Guard. 8/31/2017. Image #170831-Z-AH923-081. Downloaded 9/8/2017 from https://commons.wikimedia.org/w/index.php?curid=62096178.

Wikipedia. April 2016 United States Storm Complex. Viewed online 8/30/2017 at https://www.hcfcd.org/storm-center/tropical-storm-allison-2001.

Wikipedia. Houston. Viewed online 8/30/2017 at https://en.wikipedia.org/wiki/Houston.

Wikipedia. New Orleans. Viewed online 8/30/2017 at https://en.wikipedia.org/wiki/Houston.

California Continues to Face Future Water Supply Challenges

Despite the wet winter in 2017, climate change will pose severe challenges to California’s future water supply.

In the last post I reported that Gov. Brown has declared California’s drought emergency officially over. The state has plenty of water for the next year. This post explores the implications of this wet winter for California’s long-term water status.

I first looked at this topic in a 13-post series that ran during the summer of 2015. The series starts here. It contains a lot of information about California’s water supply and consumption. I concluded that at some point in the not-too-distant future California would experience a significant permanent water deficit. The #1 cause of the deficit would be climate change, which is projected to result in a significant reduction in the size of California’s snowcap. The #2 cause would be population increase. I performed the analysis myself because I could find no sources that did anything similar. I’m not going to repeat that analysis in this post. Rather, I’m going to report a couple of new reports that confirm the concerns I had in 2015.

Figure 1. Source: California Dept. of Water Resources.

Figure 1 illustrates the problem California faces. Almost all of California’s precipitation falls during the winter. Some of it gets temporarily “locked up” as snowpack on the Sierra Nevada mountains. Demand for water, however, peaks during the summer. California has many man-made reservoirs that release water during the summer and fall, and the state depends on the melting snowpack to recharge the man-made reservoirs as water is drawn from them. In Figure 1, the blue line represents runoff and the red line represents water demand. You can see that moving the date of maximum runoff earlier in the year increases the amount of water that cannot be captured into storage (yellow area). It has to be dumped; see the post on Oroville Dam to see what happens if the volume of water being dumped gets too high. It increases the amount of water that must be released from storage in the summer and fall. The amount released is now larger than the amount of inflow the reservoir receives, resulting in an increased water deficit (the blue area represents water received, the green area represents water discharged equal to the size of the blue area, and the red area represents the deficit). There is a water deficit in average years, but it is small, and a winter with slightly above average precipitation can make up the deficit. Moving maximum runoff earlier in the year increases the size of the deficit; now only a much wetter year can recharge the reservoirs.

Figure 2. Source: California Dept. of Water Resources.

Figure 2 includes two charts. The first chart shows the percentage of precipitation in California that occurred as rain from 1948-2012. If precipitation occurs as rain, it is not snow and can’t add to the snowpack. On the chart, the black horizontal line is the mean percentage across all years. Red columns represent years with above average percentage of rain, the blue columns below average. There is variation between years, but you can see that the red columns cluster to the right while blue columns cluster to the left. That means that on average an increasing percentage of precipitation is falling as rain. Thus, on average, unless annual precipitation undergoes a sustained increase (which hasn’t happened and is not projected), California’s snowpack will shrink, because what once was snow is now rain.

The second chart in Figure 2 shows runoff measured on the Sacramento River. The red line represents the 50-year period from 1906-1955, while the blue line represents the 52-year period from 1956-2007. This is the specific problem that was discussed conceptually in Figure 1. You can see that runoff has moved earlier in the year by about a month.


Figure 3. Source: National Centers for Environmental Information.

Why is more precipitation falling as rain rather than snow, and why is melt occurring earlier? Because of increased temperature. Winter is when the snow falls in California, and it is when the state receives the bulk of its precipitation. Figure 3 shows that the average winter temperature (December – March) has increased more than 2°F. In addition, if you look at Figure 3 carefully, you can see that the rate of temperature increase accelerated somewhere around 1980. The runoff chart in Figure 2 chunks the data into only 2 groups, each about 50 years long. Because of the acceleration in the increase in temperature, I believe that if they had chunked the data into 3 groups, each about 33 years long, the change towards earlier snowmelt would have been even greater than the one shown.

How dire is the threat is to California’s snowpack? It depends on which climate projection is used. The projected effects of climate change depend very much on how humankind responds to the threat. If we greatly reduce our GHG emissions immediately, the climate will warm less; if we don’t, it will warm more.

Figure 4. Source: California Dept. of Water Resources.

Figure 4 shows the historical size of the California snowpack plus 2 projections. The middle map show the projected size of the snow pack if warming is less. The map on the right shows its size if warming is more. You can see that, even under the low warming scenario, a loss of 48% of the snowpack is projected. Under the high warming scenario, a 65% loss of the snowpack is projected. These projections are for the end of the century. In my original series, I estimated the loss of snowpack at 40% by mid-century. That is not too far off from the high warming scenario. And I have to say, the evidence suggests that so far the world is operating under the high warming scenario, possibly, even worse.

Surface water is not the only source on which California depends. California withdraws significant amounts of water from underground aquifers, especially in (but not limited to) the agricultural areas of the Central Valley. Aquifers can be compared to underground lakes, but don’t think of them as being like a big, hollow cave in which there is a concentrated, pure body of water. Rather, think of them as regions of porous ground, such as gravel or sand. In between the pieces of gravel or sand is space, and that space can hold water. Below and on the sides are rocks or clay that are impervious to water, which allow the water to be held in the aquifer.

So long as the aquifer is charged with water, this is a situation that can last for thousands of years. If, however, water is pumped out without being replaced, then nothing occupies the spaces between the pieces of gravel or sand. If that occurs, the weight of the ground over the aquifer can compress the aquifer, reducing the amount of space available between the pieces of sand and gravel, reducing the capacity of the aquifer to hold water. When this occurs, it sometimes shows up as subsidence on the surface. In California, it is primarily the snowpack that feeds the aquifers. If a significant amount of the snowpack is lost, it will be less able to recharge the aquifers, and they will undergo increased compaction.

Figure 5. Map of Permanent Subsidence. Source: Smith et al, 2016.

As noted in my original series, significant subsidence has already occurred over California’s aquifers. More seems to be occurring every year. A recent study attempted to quantify the amount of water storage capacity being lost to compaction. The study covered the years 2007-2010, so it didn’t even include the recent severe drought (2007, 2008, and 2009 were dry years, but 2010 was 9th wettest in the record). The study covered only a small portion of the south end of the Central Valley Aquifer, yet it found that during those 4 years significant permanent subsidence had occurred (see Figure 5), resulting in a total loss of 748 million cubic meters of water storage, an amount roughly equal to 9% of the groundwater pumping that occurs in the study area. If this ratio held going forward, it would mean that for every 44.4 gallons of water pumped out each year, about 1 gallon of aquifer storage would be lost.

During the recent drought many newspaper articles reported that there had been a sharp increase in the number of wells being drilled in the Central Valley, and that the depth of the wells had also significantly increased. This suggests an increase in the rate at which the water table is being lowered, which would lead to an increased rate of compaction. As the study notes, this is a loss that cannot be replenished; aquifer storage lost to compaction is gone forever.

Dry periods become more devastating when they occur during hot periods. One reason the recent drought in California was so devastating was because it was a hot drought. A recent study found that climate change has already raised the temperature in the state (as in Figure 3 above), and will continue to raise it further, to the point that every dry year is likely to be a hot drought. The report concludes that anthropogenic warming has substantially increased the risk of severe impacts on human and natural systems, such as reduced snowpack, increased wildfire risk, acute water shortages, critical groundwater overdraft, and species extinction.

The bottom line here is that we are talking about the effects of climate change. Climate means average patterns over long periods of time – 30 years at minimum. The current wet period represents only 1 winter. Just as one swallow doesn’t make a summer, so one wet winter doesn’t make a climate trend. For that matter, neither do 5 dry years. However, California’s increase in temperature is a long-term change that does make a climate trend, and every indication suggests it will only increase more.

My conclusion is that this wet winter not withstanding, the concerns I voiced in 2015 over California’s water supply remain valid. As time passes, California will face increasing challenges meeting the demand for water (see here). The state will be unable to secure large new sources of surface water or ground water (see here), and will have to construct large, expensive desalination plants (see here). There will be sufficient water to supply human consumption if it is properly allocated (see here), but water available to agriculture will be reduced, resulting in a decline in California’s agricultural economy (see here). That loss, plus the cost of the desalination plants, will impact California’s economy (see here), as well as the food supply for the entire country.

[In the above paragraph I have referenced several of the posts in my 2015 series Drought in California. If you are interested in the topic, you should read the series sequentially, beginning with Drought in California Part 1: Introduction.]


California Department of Water Resources. 2015. California Climate Science and Data for Water Resources Management. Downloaded 4/6/2017 from http://www.water.ca.gov/climatechange/docs/CA_Climate_Science_and_Data_Final_Release_June_2015.pdf.

Diffenbaugh, Noah, Daniel Swain, and Danielle Touma. 2015. “Anthropogenic Warming Has Increased Drought Risk in California.” Proceedings of the National Academy of Sciences. Downloaded 3/30/2017 from http://www.pnas.org/content/112/13/3931.

National Centers for Environmental Information. “California, Average Temperature, December-March, 1896-2016” Graph generated and downloaded 4/13/2017 at https://www.ncdc.noaa.gov/cag/time-series/us.

Smith, R.G., R. Kinght, J. Chen, J.A. Reeves, H.A. Zebker, T. Farr, and Z. Liu. 2016. “Estimating the Permanent Loss of Groundwater Storage in the Southern San Joaquin Valley, California.” Water Resources Research, American Geophysical Union. 10.1002/2016WRO19861. Downloaded 3/30/2017 from http://onlinelibrary.wiley.com/doi/10.1002/2016WR019861/full.

Oroville Dam: When It Rains, It Pours

If you have been watching the national news, you know that California has had record precipitation for the winter so far. There has been so much rain that one of the state’s biggest reservoirs, Lake Oroville, has exceeded its capacity. Water flowing through the emergency spillway has eroded portions of the dam, threatening a dam collapse that would kill thousands and wipe out communities below the dam. Emergency efforts are underway to repair the dam.

Following are two photos that show just how badly the dam has been damaged. The photos are a bit hard to interpret, but here’s what I think they show: water overtopped the dam into the emergency spillway in amounts that the spillway was not designed to handle. In Figure 1, the water is no longer overtopping the dam, but a huge section of the dam face has been badly damaged. Several channels have been carved down the face of the dam. Figure 2 shows water surging down the damaged main spillway. The spillway is the concrete structure at left. You can see that the water isn’t flowing in it, but rather down a rogue channel that the water has cut in the face of the dam. This actually occurred 2 days after the photo in Figure 1, as the California Department of Water Resources dumped water out of the lake in anticipation of large inflows from a new storm.

Figure 1: Damage on the face of the Oroville Dam. Source: Kolke 2017.

Figure 1: Damage on the face of the Oroville Dam. Source: Kolke 2017.

Figure 2: Water Damaging the Face of Oroville Dam. Source: Grow, 2017.

Figure 2: Water Damaging the Face of Oroville Dam. Source: Grow, 2017.










It sure seems like the surface water drought in California has been broken. The most important snow survey of the year (and often the final one) occurs on or about April 1. So I’m going to wait for that date before I do an analysis of how the wet winter has changed their water situation. Hopefully, at that point there will be enough data to allow an analysis that goes beyond the headlines and looks at long-term implications.


Kolke, Dale. 2017. DK_Oro_Spillway_damage-4109_02_15_2017.jpg. California Department of Water Resources > Galleries > Dams > Oroville Dam > Oroville Spillway Damage. Downloaded 2/21/17 from http://pixel-ca-dwr.photoshelter.com/galleries/C0000OxvlgXg3yfg/G00003YCcmDTx48Y/Oroville-Spillway-Damage.

Grow, Kelly. 2017. KG_oroville_damage-1 2868_02_20_2017.jpg. California Department of Water Resources > Galleries > Dams > Oroville Dam > Oroville Spillway Damage. Downloaded 2/21/17 from http://pixel-ca-dwr.photoshelter.com/galleries/C0000OxvlgXg3yfg/G00003YCcmDTx48Y/Oroville-Spillway-Damage.