Research
Wind and solar power resource droughts: Meteorology, Climatology, and Economic Impact
Recent high-profile electric power grid blackouts in California (in August 2020) and Texas (in February 2021) are clear indications of the sensitivity of modern electricity systems – the foundation of the modern economy – to extreme weather events.
Historically, human systems have been becoming less and less affected by environmental variation but three broad trends make it so that our electric power systems – both their supply of electricity and their demand – are being impacted progressively more by the environment.
The first trend is that due both to legal mandates as well as rapid cost declines, wind, and solar power are increasing their fraction of total electricity generation. Because wind and solar power resource availability are intrinsically tied to the weather, the buildout of these forms of generation implies a system that is much more sensitive to weather variation than it has been previously.
The second broad trend is that increasing greenhouse gas concentrations will continue to increase their impact on electric power supply and demand. There is no doubt that increasingly intense heatwaves will cause unprecedented electricity demand peaks concurrent with stress on supply (solar panels, wind turbines, and transmission lines are all less efficient under extreme heat). There is also active research on whether cold air outbreaks may be expected to become more frequent under certain conditions which would also stress both electricity supply and demand. Lastly, increasing greenhouse gas concentrations will affect the mean, variability, seasonality, and geographic concentration of the wind resource, the solar resource, their covariability with each other, and their covariability with temperature.
Finally, climate and sustainability goals call for progressively more of our energy infrastructure (e.g., transportation, ranges, space, and water heating) to be electrified which would effectively amplify the societal impact of weather influences on electricity generation.
The situation can be summarized as 1) progressively more and more of society’s energy consumption is coming via electricity 2) progressively more and more of society’s electricity is being generated from weather-dependent resources like wind and solar power and 3) the mean and variability of the weather feeding this energy system is changing.
There is much literature on annual mean solar and wind power resources throughout the world, as well as heating degree and cooling degree days. There is also emerging literature on the general topic of the (co)variability and complementarity of wind and solar power resources. However, a major gap in the current state of knowledge is that there has been very little research to date on wind and solar droughts (and demand floods) from an extreme event and compound extreme event perspective (Figure 1).
In the Weather, Climate, and Human Systems lab at San Jose State University we investigate the meteorology, climatology, and economic impact of wind, solar, and temperature variability.
Below is a background video on some of the challenges associated with an electricity system that is heavily reliant on wind and solar power.
For more details on the concepts above see the National Renewable Energy Laboratory’s Western Wind and Solar Integration Study.
Below is output from a weather model (GFS), over an arbitrary 2 week period, showing the evolution of the jet stream (upper left), clouds (upper right), temperature (lower left) and near surface wind (lower right), giving an idea of how the wind, sun and temperature covary within the context of synoptic-scale weather patterns.
Comprehensive assessment of variability from hourly to millennial timescales. What are the likelihoods of wind and solar droughts of given magnitudes? To know what level of risk we are willing to tolerate, we need the best possible assessments of return periods on long timescales. For example, is a given drought a once per year event, a once per decade event, a once per century event, or a once per millennium event (and how are these shifting over time)? We use the best available reanalysis and simulations of past and future climate to answer these questions. Below are some illustrations of analysis assessing variability from the hourly to millennial timescale using WRF reanalysis from the years 1990 to 2018 and the MRI-ESM2 climate model from the year 850 to 2100 (SSP2-45 from 2015-2100).
Meteorology and Climatology of Wind and Solar Droughts. Below are some plots illustrating the seasonal cycle in available wind power and solar power as well as heating and cooling degree days over western North America. For more details, see the preprint here: https://www.researchsquare.com/article/rs-433450/v1
Climatology of potential power supplied by wind and solar resources and a proxy for power demanded via heating and cooling degree days. Left) spatial distribution of climatology for different portions of the year. Right) temporal distribution of climatology with 1st percentile drought weeks highlighted.
Every wind and solar weekly value over western North America between 1979 and 2018
Below are some plots showing the typical synoptic set up of the atmosphere, as well as sea surface temperatures, during droughts in wind and solar resources over western North America.
Composites over the 20 weeks with the lowest domain-average wind power (a,d,g,j), solar power (b,e,h,k) and combined wind+solar power (c,f,I,l). The variables displayed are described on the left. All anomalies are defined with respect to the typical value for that week of the year and the stippling shows where at least 15 of the 20 weeks (75%) showed anomalies of the same sign.
Climate Change: We are investigating how extremes in wind resources, solar resources and heating/cooling demand are changing as greenhouse gas concentrations increase. Below are annual trends (1950-2020) fit to the most extreme-stress week of each year.
Left) Historical trends (at the dot in the right panel) of the single week of the year with the highest-stress conditions (lowest wind and solar power week or highest heating degree or cooling degree days). Right) Map of trends in the variable labelled on the left. The wind resource is converted to a realistic spatial density using a wind power curve and an assumption about turbine spacing while the solar resource is left as the raw incident solar radiation.
Global Teleconnections: We are investigating how modes of variability like the El-Nino Southern Oscillation (ENSO) are ‘teleconnected’ to variability in wind and solar resources globally. Since ENSO has some predictability on seasonal timescales, this research has the potential to open the door to seasonal wind and solar resource forecasts.
Top) Mean wind speed anomaly when NINO3.4 SSTs are > +1 sigma. Middle) Mean wind speed anomaly when NINO3.4 SSTs are < -1 sigma. Bottom) Top minus middle.
High spatiotemporal resolution analysis. Below is an illustration of an assessment of California wind, solar and degree-people-hours at 2kmX2km spatial resolution, 1 hour temporal resolution and over 30 years (1990-2019). The likelihood and causes of very high stress hours are assessed.
Depiction of research combining geolocated population, wind farms, solar farms and high spatiotemporal resolution (2km and 1 hour) historical weather to assess the highest stress situations on the electric power system.
Some current projects involve:
Diagnosing the synoptic meteorological situations that lead to extreme events in the magnitude of windiness, surface solar radiation and temperature in different regions.
Investigating teleconnections between extreme events in the above three phenomena and large-scale states of the climate system (e.g., the El-Niño Southern Oscillation, the North Atlantic Oscillation, etc.).
Investigating long-term trends in the occurrence of these events and projections of these phenomena in the coming century.
Diagnosing the reasons for uncertainty in projections of these phenomena in state of the art global climate models.
Quantifying the costs of these extreme events and the economic value of forecasting these events ahead of time.
Our research will provide a foundation for planning a renewable energy system that will be robust to the most extreme stresses that weather events have to offer both today and in the future. If you are a motivated undergraduate student studying mathematics, computer science, engineering, meteorology, ocean science, earth science, environmental science or a related field and are interested in gaining valuable skills and pursuing fun and impactful research, please contact me, regarding the pursuit of a master’s degree in our lab.
Climate Change and Fire Activity
We also work in close conjunction with our colleagues in the Fire Weather Research Laboratory to study how humans influence wildfires both directly as well as indirectly through the impact of increasing atmospheric greenhouse gas concentrations.
We are currently conducting research that will help Pacific Gas and Electric (PG&E) better understand the fire weather conditions associated with extreme wildfire behavior. This research will help inform their policies regarding Public Safety Power Shutoffs (PSPSs) which affect millions of Californians.
Specifically, we are conducting analyses using PG&E’s new 30-year climatology of 2 km WRF model output (1989-2019). These data will allow for the first analyses of critical fire weather conditions using an unprecedented combination of high spatiotemporal resolution and long duration.
Individual projects being undertaken include analysis of:
Climatology and decadal trends in fire weather and Diablo Wind events, or other Foehn wind events (type, intensity, duration, etc.).
Covariation of fire weather mesoscale circulation patterns with synoptic patterns and known modes of variability like ENSO, PNA, MJO, etc.
Seasonal to sub-seasonal predictability of fire weather circulation patterns using various antecedent predictors (SSTs, arctic sea ice, MJO phase, etc.) and machine learning algorithms.
Climatology and decadal trends in ignition potential and high fire danger conditions from convection-inferred dry-lightning events.
High-resolution trends in existing fire-weather indices and local fire season duration.
The development of a new PSPS index for extreme fire weather events that takes into account utility circuit stability, load.
The relative influence of changes in aridity vs. changes in atmospheric circulation on fire activity including nocturnal drying events.
The influence of changes in wet-season precipitation on fire danger and burned area via its influence on fuel loads and live fuel moisture content.
The relative influence of the expansion of the WUI and transmission infrastructure on fire activity vs. climatological factors.
Empirical Temperature Forecasting
We conduct research on machine learning methods for forecasting various environmental variables. See here for a global temperature forecast from an empirical method for predicting year‐to‐year global temperature progression (full study: Brown and Caldeira, 2020). See a research poster presented at the American Geophysical Union’s fall meeting and a research talk at the American Meteorological Society’s annual meeting for potential future applications of this research.
Climate Change Mitigation Cost-Benefit Analysis
We conduct research using wholistic integrated assessment models to assess when the investment in decarbonization begins to pay off. See also a research poster, paper and interview at the American Geophysical Fall meeting.