California is one of a handful of Mediterranean climates in the world, characterized by hot, dry summers and cool, wet winters. The state depends on a few big winter storms to provide the runoff needed to satisfy water demands in the summer and fall. However, the last three years, fewer storms have traveled through, leading to severe drought, and some records being broken. Is this a sign of climate change?
On July 25th, the Delta Science Program, Ecosystem Restoration Program & Surface Water Ambient Monitoring
Program hosted a Brown Bag seminar featuring Michael Anderson, State Climatologist for the Department of Water Resources who discussed the current conditions, placing them in the context of decadal variation and longer historical climate records.
“California’s topography is one of the drivers of our weather and climate,” began Michael Anderson. “All those fantastic mountains get in the way of moisture flowing into the state from the Pacific Ocean, and that creates our rainfall.”
“Atmospheric river events are important to California, both for water supply and for flooding,” he said, presenting a slide depicting an atmospheric event. “A whole number of processes going on in the Pacific Ocean need to line up in a particular way and when they line up in just the right way, we can have a flood,. From a drought perspective, there are a number of the ways things can miss, and it doesn’t too many misses in the 90-day window where we get half of our annual precipitation for things to fall apart.”
He explained that the colors on the image indicate water vapor, with the reds, oranges and purples being high concentrations of water vapor; the yellow and green have less and the blue is dry air. “That’s the moisture element; then we need wind and a storm system to run into that plume of moisture, and that give us our heavy precipitation.”
He then presented showing water vapor totals at various monitoring stations from December 27, 2013, noting that this is data from the new extreme precipitation monitoring network that was installed over the last five years. “What these little numbers on the map are showing is the amount of water vapor that’s in the air over California, so if you took all the water from the surface, all the way to space, and condensed it out, that’s how many centimeters of water would be there. This is in December, one of our wettest months, and in Sacramento, there was all of half of centimeter of moisture; Yosemite had half a millimeter. It’s not going to rain when things are this dry. If you don’t have moisture, we’re not going to get our rain.”
He then presented a slide of a precipitation event, explaining that anything over 2 ½ centimeters, or about an inch of moisture, can start producing precipitation. “Most of the moisture in the atmosphere is in first 10,000 feet, so it’s coming off the ocean and it in runs into mountains, there is combined convergence and lift, and that benefits our precipitation. You can see down there in the deserts, not as much, get up into Oregon and not as much. These are fairly narrow bands of moisture.”
He then presented a slide of the Northern Sierra 8 station index, noting that the eight stations extend from Mount Shasta City down to Pacific House, which is the area where most of our water supply originates. “This is a precipitation index with an average of 50 inches. Our wettest year at the top was 88.5” in an El Nino year, 82-83, and one of the driest years was 76-77, where a blast in September knocked it off from 23-24 being the driest, but both of those were pretty miserable years.”
“Funny thing about 76-77, it was an El Nino year,” said Mr. Anderson. “So now when people tell you it’s going to be an El Nino year, you know that’s it’s been the driest in the last 60 years and it’s been the wettest in the last 60 years, so you should have some good information there.”
He pointed out that the other green line, which represents 2012-13, shows a steep jump an atmospheric river event in December. “We had high water in the Russian River, and at the time we passed 82-83 and we were on track for being wetter than the wettest year on record. Around Christmas, another blast, and we’re thinking to ourselves, this is great. We picked up a third of our water year precipitation in this event … What could go wrong? Well, look at the next 14 months. Not much happens … by the beginning of February, you have crossed the entire distribution of historical water years in California in a 14 month period. We went from wetter than the wettest and after 14 months of absolutely nothing happening, now we’re drier than the driest and things are looking a little uncomfortable.” He noted that some precipitation in February and March put us above the driest year.
“Those atmospheric rivers are important elements of our precipitation,” he said. “This one that got us out of the worst was about 8” in two weeks. Another was about 14” in 8 days. So there is a difference to them, and what size you get, but we still need them to get the supply we need.”
He then presented a slide of the San Joaquin 5 station index, noting that 2012-13 is in purple on this graph. “You see that big step at the beginning of December didn’t get as far in terms of getting up to the wettest end of the distribution, and it still goes flat as well, and it stays flat into the royal blue line that digs under the driest year. Down here it’s been drier than what’s up in the Sierra … up north, as bad as it is, its still doing better than as you head south.”
He then presented a slide plotting snow water content for the last three years against the wettest and the driest years. He noted that in 2011-12, at one point the northern snowpack got up to average at one point but even that didn’t last. “And in the central and south, they only got to 50%, and then for the next two years, hanging out at that about 50%. You do three years of only a 50% snowpack – and remember 100% snowpack does not equal 100% runoff; you’ve got dry conditions interceding, and our largest reservoir that we rely on for our water supply is not quite holding up. Mother Nature didn’t deliver.”
“Here’s a different way to look at it,” he said, presenting a slide, explaining that JPL is doing an experimental product to take MODIS data and create a model estimating snow water equivalent. The image on the right is the elevation band, and the image on the left is the model average over 13 years of snow water equivalent for May 4th.
He then presented a slide showing the snow water equivalent for May 4th of this year, as well as April 27. “Particularly at the front of the basin, there isn’t a whole lot of snow showing up … not nearly what you’d want for this time of year.”
Mr. Anderson said a colleague of his went through CDEC, taking summary data for 154 reservoirs, grabbing each month’s percent of average since 2010 and plotted it relative to average. “What we’ll really focus on here is in 2012, beginning in January, the big storms were well above average, 120%. doing great, but things are dry … we start trailing down and 2012 we end up below average. … We get a little bit and then look at the dive in 2013. When you go extremely dry and you have been wet, and then get into December … you’re rivaling your wettest year, what are your expectations for your water supply and environmental needs? You’re going to think we’re all set for runoff, well if it doesn’t happen, then you’re going to borrow storage to do it. We get into 2014 and it’s still tailing down, as a percent of average, you can still say though for as awful as it’s been, we’re still faring better than we did back in 77.”
He then presented a slide for Shasta Reservoir. “Here in 2012-13, the black line is above the light blue which says we’re above average until we start delivering water, and we start digging down into our storage … we don’t get to 77, then snowmelt kicks in and then we’re digging down, and currently, as of midnight last night, we have a little less than 1.4 MAF in Shasta. For this time of year, we’d like it to be over 2.5 MAF.”
“Here’s the story of Oroville, somewhat similar,” he said, presenting a slide for Lake Oroville. “When we dug down, what did we manage to do – we dug under that 77 line and we dug into it for quite a little ways until spring melt. Those February-March storms kicked in to help out … now we’re down to 1.18 MAF and we want to be above 2.25 MAF, so this is a lot of incidental flood space storage. There’s always a bright side.”
He then presented a slide of Folsom Reservoir and noted that at the beginning of February this year, the reservoir was well below the 1977 line. “Then in February, 100,000 acre-feet of storage comes in and that takes an absolutely miserable situation and makes it better than we were, and then tailing down, relative to the other years and lines.”
It’s a tougher story at San Luis Reservoir, he said. “The line in 2012 is hanging out with the driest year, and then heading into this year, well under it, almost the whole year, and towards June, finally snuck above the worst year. Now you get to ask yourself, where’s the new worst year. Possibly a new line to add to the chart.”
“That’s where we are,” said Mr. Anderson. “Now for a little context.”
He then presented a scatterplot of average mean temperatures plotted against average mean precipitation for the Sacramento River Basin from 1895 to 2013. He explained that the blue diamonds are the average temperature for each year and annual precipitation for each calendar year, the big triangle is the period of record average, and the black squares are the 21st century. “When I look at this, it seems to me we’re sampling preferentially in the distribution on a particular side. This is calendar year data so this is a little different than how water managers normally look at the world. It’s a different perspective.”
“But look at 2013,” he pointed out. “We found a new part of the distribution that hadn’t been mapped before. When people tell you this is the worst drought since 76-77, 76 was the previous low point on that chart in terms of precipitation. Look at the difference in this year – it’s two degrees warmer, so there are additional challenges to be had 40 years later.”
“The story isn’t different when you move into the San Joaquin Basin,” he said, presenting the same plot for the San Joaquin River Basin. “Again 2013 was warmer and drier than anything that had been seen in 100 years previous. The 21st century is sampling distinctly on one side of the distribution.”
Statewide, same story again, so calendar year 2013 was definitely one that you can say we hadn’t seen before.”
“Now we’ll take it from an even different perspective,” he said. “This is 36 month accumulation statewide, climate division data, but this time we’re parsing by precipitation year, so July 1st to June 30th. 1981-84 which includes that 82-83 El Nino year was our wettest one there with 94.78” of precipitation in this data set. 1974-77 used to be the lowest when we totaled 46.65; the average should be close to 70 for that three year accumulation. When we look at the past three years, we’re down to 45.19, so from this perspective over that three year period, we’ve plumbed a new low in our precipitation with the climate division data set at NOAA.”
He then presented a slide of the 8-station index cumulative deviation, noting that he produced this graph by taking the shortfall from average each year, as long as the year before was below average, and appended that year onto the previous year shortage. “So you see these big long six year droughts. 1929-34 and 1976-77 where in 2 years you really dig deep, 1924 the single driest year, and then 2014 -you see how this year matches up from that perspective. And as it says, it assumes no further precipitation this year.”
He then ran down how this year shapes up in the rankings. At 29”, the 8-station index is the 8th driest water year, and the 5 station index is the third driest. The 8-river index, the runoff from the eight rivers in the Central Valley, it’s the fourth driest.
Mr. Anderson said that in the 20th century, there were 2,3,4 and 6 year droughts so we’re not seeing anything out of the usual yet. “The average precipitation in the multi-decadal dry period from 1910 to 1940 is 20% less than the period of record average, so that makes things a little more challenging to manage too, if Mother Nature on average for quite an extended period of time isn’t delivering.”
The paleorecord shows multiple 10+ year droughts as well as 2 century-long dry periods. “One of the outcomes of the drought was that the University of Arizona refined the paleo-hydrology study for the Sacramento, San-Joaquin and Klamath Basins,” he said, presenting a slide of some of the results. “In the Sacramento Basin, there are these long periods where things kind of go dry, and San Joaquin and then the Colorado River, and we see those decadal-scale wiggles.”
In terms of temporal variability, California’s inter-annual variability is greater than anywhere else in the continental U.S, he said, presenting a map depicting the standard deviation of annual precipitation from the mean annual precipitation for the nation. “Everywhere east of the Mississippi River, their standard deviation relative to their mean annual precipitation is between 10 and 20 percent,” he said. “One year looks an awful lot like the next. … Here, down here in the deserts, the mean is small, and year to year variation can be 70% of the mean, but notice that on the North Coast the mean is 120 inches of precipitation, and you still have a variability that’s 40% of that. That’s just downright jaw dropping. We can say we do things differently out here.”
He then presented a graph titled ‘Decadal Scale Variability.’ “Here’s a bunch of different data sets with ten year moving average put through it, and entertainingly enough, the same pattern kept falling out,” he said. “I put LA County down there at the bottom just to show Southern California, because some of these patterns extend across the state. The other four are the 8-station and 5-station index and Sonoma County and Mendocino County. It’s fascinating that the two counties that sit right on top of each other have patterns and magnitudes that are very similar to the 8-station and 5-station index. But you see even the little wiggles line up really well.”
“So let’s dig into the 8 station index and some of the entertaining things that have been happening with it,” he said, presenting a slide of the Northern Sierra 8-station index from 1921 to present day. “We see the big swings year to year. Nine of the last fourteen years have been below average. The 21st century so far hasn’t quite been the same. If you break it down and look at the period of record average, from 1971-2000, that average was 53 inches. We were rolling, we had the biggest year and also the lowest year in the last 60, but then we hit the 21st century, and those 14 years are not quite up to that period of record average.”
He then presented a slide showing the monthly breakdown of the 8-station index, noting that in December, there’s a big spike in the precipitation totals for the 21st century. “2012 and 2014 were the third and fourth driest Decembers in 90 or so years of record … How can you have a big spike in 14 years of average when you are including some of the bottom feeders there in the past three? You have some relatively wet Decembers where you have over 20 inches that kind of skew those statistics a bit. “
He next presented a slide for the Sacramento River, noting that the period of record average is 18 MAF, and almost 19 MAF from 1971 to 2000, but in the 21st century, it’s down to 16 MAF. “So if 18 MAF is the long term average, you’re down 2 MAF from that; but if you’re trying to deal with how you’re managing things at the close of the 20th century, you’re working 3 million less, and it’s a bit of a trick to do.”
Things are dry in the south as well, with ten of the last fourteen years on the 5-station index below average, he said. “You see here again, closing out the 20th century that things were a little wetter than the period of record average, but the 21st century has definitely so far been shorter on that.”
“We tied with the driest water year in 2012 when no precipitation fell in December of 2011, and December of 2014 was 10th driest with only 1.1 inches,” he said. “Now if you’re looking at period of record average in December of over 6” and the 10th driest is only an inch, there could be some really dry years down there in the San Joaquin offset by some really big years, getting back to the notion that atmospheric rivers are really important to how things shape out for us.”
He then presented a chart of the monthly distribution of San Joaquin River runoff, noting that it’s an entirely different shape because the mountains are tall and things are driven by snowmelt. “We see that the period of record is a little below 6 MAF, from 1971 to 2000 a little above 6 MAF, and in the 21st century, only 5.3 MAF of runoff. So, again we’ve got 14 years now where we’re working with a deck that isn’t fully loaded, so to speak.”
This often brings the question, is this climate change, he said. “In climate division data, we’re sampling distinctly on the warm side of the distribution, and things that historically were at the edge, we’re now giving them a lot of friends to work with. Those that were left in the period of record average, we don’t seem to get back there anymore.”
We hear the expectations: A smaller snowpack, more rain, less snow, earlier snowmelt onset, more variability, and more extremes, and the next question that pops up is are we seeing these already, he said. “Well, yes … but how you define them and how they manifest themselves and how it consistently behaves over years and years – not just one oddity year, because remember year to year, we have the most crazy variability.”
“This is the fun thing when people talk about trends,” he said. “Do you know how big the trend has to be to show up with that kind of variability? That’d be a little scary, so we have to tease things out a little more. The Western Region Climate Center set out to try and tease this out is through the California Climate Tracker,” he said, presenting a slide showing historical precipitation and explaining that the climate tracker is a tool available online that allows you to look at datasets going back to 1895. “For precipitation for our water year to date, 2014, if you see red, that means it’s never been drier, and if you see orange, that means you’re in the top 10, and if you see yellow, that means you’re still below average but not doing so bad. We had a pretty good monsoon for the Colorado desert, but even that didn’t’ quite bail things out.”
“In terms of temperature, there’s a lot of red particularly along the coast and particularly where you have a lot of California’s population, particularly in those maximum temperatures,” he said. “About the only place that isn’t cooking, so to speak, is up on the north coast where things are a bit cooler but still not below average; they are still either average or above.”
He then presented a time series for temperature departure for both maximum and minimum temperatures. “You start to see particularly on the right with the minimum temperatures that it’s getting to where you don’t see too many blue lines, which is where the departure is below the 1949-2005 average, so we’re starting to sample consistently on one side of the checkpoint.”
The Department of Water Resources is working on a new metric describing rain as a percentage of total precipitation, he said. “Here’s their first shot at it; they are using some graded data products for temperature and precipitation to try and tease this out. The mean is for the region of the state where snow is important. … There are a lot of different elements in a water year that go into whether you’re getting more rain or snow, including total amount of precipitation or a big storm that’s either warm and wet or cold. It’s not as if things are monotonically increasing, and there is a lot of variability to sift through, but you start seeing that you’re sampling preferentially.”
So, to summarize, the current drought along with 21st century droughts have shown record setting characteristics and are warmer than 20th century counterparts, he said. “Instead of just saying it hasn’t rained as much, you really have to start bringing in temperature as an element to describe how these droughts are rolling.”
Atmospheric river events provide significant inputs into annual precipitation totals, but we’re seeing fewer such events in drought years, he said. “If you look at the water year outcomes, you see some clustering and it relates to the number and size of the atmospheric river events that do make their way through, so it may seem odd, but studying the wet extremes can help inform you on the dry side as well.”
The expectation of increased variability means new extremes and extreme transitions. “2013 is going to be a pretty good example of where we started at the wettest end and by the time we get into 2014 we’re on driest of driest,” he said.
Planning for future droughts can take advantage of information in the historical record including paleo reconstructions. “Don’t think just because the future may not be reflective of the past that you have to abandon everything you’ve seen in the past,” he said. “We have a lot of variability; there’s a lot of elements in there. They are waiting for you, you have to do a lot of digging, but there’s a lot of good information there. The trick here is to increase your understanding of the causal mechanisms and watershed condition response – what happens when the watershed is wet, and as it dries out, and then when it is dry, how much effort does it take to wet it back up to where you’re getting runoff you can refill the reservoirs with.”
“And with that … “
What do we know about the effect of warmer temperatures on atmospheric rivers?
“Fantastic question, because a warmer atmosphere can hold more water,” Mr. Anderson responded. “The part we’re still working on is digging through history and finding proxy-measurements. We have a new network to directly observe these but its only 5 years old, so not a whole lot of data to go in with that. We’re looking at proxies we can get from the historic record and look through and see which elements of them are getting bigger and more water vapor flux change in the duration of the event, which may lead to a bigger storm, but if you don’t get enough storms or you start from a drier condition – maybe if you do get one storm, you get a bigger bang for your event, but we’re still digging into it.”