Transcript Document

How bad is climate change going to impact water delivery?
Known facts about climate change…
•
•
•
Average global surface temperature has increased by 0.6°C since the late 19th century.
Evidence of reduction in Artic sea ice and continental glaciers.
1.0 to 2.0 mm/yr of global average sea level rise during the 20th century.
Some apparent reasons for relatively small water
delivery impacts simulated…
•
Use of historical hydrology
 Climate models project long term trends and may not account for all critical factors
influencing climate variability (year-to-year and month-to-month).
 Most of the studies utilize a climate change forecast superimposed on the 20th
century hydrologic sequence. This presents no new information about potential
climate variability and more critical hydrological sequences.
Likely reductions in California water supply…
Most global climate models estimate an average increase of at least 2oC over the next
century under business-as-usual scenario. This could:
Reduce spring snowmelt  increase need for water storage in reservoirs
Increase urban water use, increase evaporation and evapotranspiration
Etc.
•
•
•
•
Smart flood control
 Models (e.g. EBMUDSIM) know when to begin reservoir refill according to how
much snowpack remains in the watershed.
 Depending on the system’s rainflood requirements, there is enough snowmelt to refill
in all but a few years of the historical record due to a 28% shift of runoff.
 If those years were to precede a critical drought then results could be more
substantial… The sequence of wet and dry years is important!
But by how much?
Some quantitative results on potential water
delivery impacts in California…
System carryover storage (TAF)
•
System carryover storage under future level of development
Historical hydrology
600
Warming (i.e. early snowmelt)
 Preliminary studies using CALSIM II, Bardini et.al. 2001
– 20th century diminishing snowmelt trend extended
– 2.5% reduction in total CVP/SWP exports during 1928-34 drought
 Preliminary studies using EBMUDSIM, Richards 2003
– Shift 28% of historical Apr-Jul runoff to Nov-Mar
– <1% reduction in Mokelumne deliveries during 1928-34 drought
Difference due to climate change scenario
500
400
300
200
100
0
-100
1921
1928
1935
1942
1949
1956
1963
1970
1977
1984
1991
1998
Reduced precipitation
 Brekke et.al. 2003
– PCM scenario w/ 13% reduction in San Joaquin River region annual runoff
– Negligible reduction in dry-year deliveries to east-side contractors; 14% reduction in dryyear deliveries to west-side contractors (Delta exporters)
 Richards 2003
– 10%, 20%, 30% reduction in annual runoff
– 4%, 6%, 12% reduction in dry-year deliveries to Mokelumne water users
Drought Avg
What are some alternate approaches?
Focus on critical factors affecting water delivery…
•
200
150
•
100
50
0
25
50
75
100
Climate research and modeling:
 Climate variability – Quantify linkage between precipitation and major drivers (e.g.
PDO, ENSO); identify factors influencing extreme events (e.g. “Pineapple Express”)
and quantify probability of occurrence in each watershed.
 Watershed hydrology – Quantify relationships between climate variables and
hydrology variables (e.g. temperature and snowpack, precipitation and runoff) in
each watershed.
Long term Avg.
250
Annual Delivery (TAF)
•
125
% of Historical Runoff (1921-2000)
Not many, and by not as much. Why?
Water delivery impacts research and modeling:
 Reservoir operations – Quantify potential changes in hydrology-based rules and
agreements under climate changes (e.g. flood control, target storage, water use,
instream flows). Develop models to relate project operations to hydrology rather
than utilizing historical data and empirical formulations.
 Mitigation strategies - Lund et. al., 2003 and VanRheenan et.al., 2004 have explored
mitigation strategies to cope with climate change impacts. This is an important
aspect of understanding the problem.
 Conduct more comprehensive sensitivity analyses.
…and break the problem into smaller pieces.
Kevin Richards and K.T.Shum, EBMUD - California Water and Environmental Modeling Forum Annual Meeting, 2004
Known facts about climate change…
•
•
•
Average global surface temperature has increased by 0.6°C since the late 19th century.
Evidence of reduction in Artic sea ice and continental glaciers.
1.0 to 2.0 mm/yr of global average sea level rise during the 20th century.
Likely reductions in California water supply…
Most global climate models estimate an average increase of at least 2oC over the next
century under business-as-usual scenario. This could:
•
•
•
Reduce spring snowmelt  increase need for water storage in reservoirs
Increase urban water use, increase evaporation and evapotranspiration
Etc.
But by how much?
Some quantitative results on potential water
delivery impacts in California…
•
Warming (i.e. early snowmelt)
 Preliminary studies using CALSIM II, Bardini et.al. 2001
– 20th century diminishing snowmelt trend extended
– 2.5% reduction in total CVP/SWP exports during 1928-34 drought
 Preliminary studies using EBMUDSIM, Richards 2003
– Shift 28% of historical Apr-Jul runoff to Nov-Mar
– <1% reduction in Mokelumne deliveries during 1928-34 drought
•
Reduced precipitation
 Brekke et.al. 2003
– PCM scenario w/ 13% reduction in San Joaquin River region annual runoff
– Negligible reduction in dry-year deliveries to east-side contractors; 14% reduction in dryyear deliveries to west-side contractors (Delta exporters)
 Richards 2003
– 10%, 20%, 30% reduction in annual runoff
– 4%, 6%, 12% reduction in dry-year deliveries to Mokelumne water users
Not many, and by not as much. Why?
Some apparent reasons for relatively small water
delivery impacts simulated…
•
Use of historical hydrology
 Climate models project long term trends and may not account for all critical factors
influencing climate variability (year-to-year and month-to-month).
 Most of the studies utilize a climate change forecast superimposed on the 20th
century hydrologic sequence. This presents no new information about potential
climate variability and more critical hydrological sequences.
•
Smart flood control
 Models (e.g. EBMUDSIM) know when to begin reservoir refill according to how
much snowpack remains in the watershed.
 Depending on the system’s rainflood requirements, there is enough snowmelt to refill
in all but a few years of the historical record due to a 28% shift of runoff.
 If those years were to precede a critical drought then results could be more
substantial… The sequence of wet and dry years is important!
What are some alternate approaches?
Focus on critical factors affecting water delivery…
•
Climate research and modeling:
 Climate variability – Quantify linkage between precipitation and major drivers (e.g.
PDO, ENSO); identify factors influencing extreme events (e.g. “Pineapple Express”)
and quantify probability of occurrence in each watershed.
 Watershed hydrology – Quantify relationships between climate variables and
hydrology variables (e.g. temperature and snowpack, precipitation and runoff) in
each watershed.
•
Water delivery impacts research and modeling:
 Reservoir operations – Quantify potential changes in hydrology-based rules and
agreements under climate changes (e.g. flood control, target storage, water use,
instream flows). Develop models to relate project operations to hydrology rather
than utilizing historical data and empirical formulations.
 Mitigation strategies - Lund et. al., 2003 and VanRheenan et.al., 2004 have explored
mitigation strategies to cope with climate change impacts. This is an important
aspect of understanding the problem.
 Conduct more comprehensive sensitivity analyses.
…and break the problem into smaller pieces.
Preliminary results from EBMUDSIM
Drought Avg
Long term Avg.
250
Annual Delivery (TAF)
200
150
100
50
0
25
50
75
100
125
% of Historical Runoff (1921-2000)
Kevin Richards, EBMUD