GRACE can resolve nearly uncorrelated mascons that are blocks 400km on each side with a noise floor of ~1cm equivalent water height. (This is latitude dependent because GRACE's denser ground tracks near the poles allow for better resolution.) Each mascon has a mass of ~1.6 gigatons, and a fully-loaded coal train is ~10 kilotons, so GRACE falls short by about five orders of magnitude.
The improved laser ranging on the GRACE follow-on will increase sensitivity, and David Wiese analyzes improvements due to lowering the satellites' altitude and/or adding more satellites to the GRACE system.
You're right to suspect that detecting a tiny change in local gravity is limited by uncertainties in models such as atmosphere dynamics. I've discussed how GPS occultation data (among many other data sources) can be used to reduce these uncertainties.
Other anthropogenic effects such as groundwater depletion can already be detected with GRACE. Rodell et al. 2009 (PDF) and Tiwari et al. 2009 (PDF) observed this in northern India, and Famiglietti et al. 2011 (PDF) recently observed similar groundwater depletion in California.
Also, Barton Paul Levenson explains the World Meteorological Organization (WMO) definition of climate, which uses a more scientifically conservative 30 year minimum timespan.
CO2 levels parallel average global temperatures, in looking at the geological record, correct? That gives much credence to CO2 as a mechanism for current climate change.
Correct. Here’s a figure from Royer et al. 2007 which concludes that “a climate sensitivity greater than 1.5C has probably been a robust feature of the Earth’s climate system over the past 420 million years”.
What increased C02 in ancient times? It is also my understanding that volcanic action produced CO2 which raised temps, Correct? Also, at other epochs, certain orbital factors cooked CO2 out of oceans. In this latter case, I would expect CO2 levels to lag temps (and then force greater changes) and in the former (volcanic) case CO2 levels would precede temp rise. Correct?
Correct. I've discussed the difference between these situations.
Here is a related question: If the former case, volcanoes produced CO2 and that raised temps, that same rise should also cook CO2 out of the oceans, which would produce an even greater rise in temps. What could reverse this trend?
That's a good question; I described the difference between climate and weather at the beginning of my article. I later updated it with a better analogy from NOAA: One way to distinguish between weather and climate is that the climate of your hometown will determine how many sweaters you have in your closet. The weather will determine if you should be wearing a sweater right now.
Many times the climate being discussed is global, so an average is taken over the entire Earth. For global temperatures, Santer et al. 2011 shows that one needs to average over ~17 years of data to obtain statistically significant climate trends. Here's another explanation by Tamino. Also, the Skeptical Science trend calculator helps visualize statistical significance.
What in the world was up with this letter from ex-employees [plantsneedco2.org] (also discussed on Slashdot [slashdot.org])? Was that just totally out of left field? Was there an internal reaction to it? Did you respond?
Yes, it was totally out of left field. I responded in an internal JPL email, and copied the email to my (other) website.
-Bryan Killett, aka khayman80, aka Dumb Scientist
(This was copied from very far down so readers don't have to wade through all these charming comments just to read my answer to eldavojohn's question.)
For those who may not be aware, GPS signals can be used to measure atmospheric properties. As GRACE (or any satellite with a modern GPS receiver) listens to a GPS satellite that's about to pass below the horizon, the GPS signal passes through the atmosphere. Thus the GPS signal is refracted and delayed in ways that can reveal the temperature, pressure, and refractivity of the atmosphere at different altitudes. These are known as GPS occultation measurements.
I've never used GPS occultation measurements, so I asked Gerhard Kruizinga about them at the weekly GRACE meeting. He pointed me to GFZ which has near-real time data (that page shows GPS occultation measurements taken today, but getting the data from GFZ probably requires filling out a quick form). Gerhard also mentioned that GPS occultation measurements from GRACE are regularly fed into the ECMWF atmosphere model, which is briefly described here.
GRACE also measured the 2005 Amazon drought, regarded as the worst in over a century. Just five years later, the 2010 Amazon drought might have been even more severe.
GRACE also measured the 2010-2011 floods in Australian and Columbia, which dumped so much water on land that sea level temporarily dropped by ~6mm.
Does warmer water in the summer weigh more or take up more volume?
Warmer water doesn't weigh more (except for absurdly small relativistic corrections), but it does take up more volume. This is called the thermosteric effect;...
There's no requirement to do so. For instance, ask for the code to the Hubble or WMAP telescopes, or most of the other code produced at NASA. You'll probably be disappointed. I personally think that government funded science missions should release their code in some manner, but I don't want to start a license war. Personally, I chose a license rather than releasing my code to the public domain because I was nervous that a corporation might "invent" my code and sue me for using it. If you'd like different terms than the GPLv3 provides, please contact me and I'll oblige for reasonable modifications.
I've previously discussed some of the advantages of GRACE 2, which is the real successor to GRACE and has a vague launch date in the (hopefully early) 2020s. It'd be great if we could find some way to service and refuel these satellites to extend their lifespans, especially if the low-altitude drag-free option is chosen. But for now we're trying to do what we can with less funding.
GRACE is dying. Its batteries aren't reliable enough to keep the satellites' microwave ranging system running continuously now. The JPL engineering team has worked miracles to give us 10 years of GRACE data, especially since the mission was only supposed to last 5 years. The GRACE follow-on mission is scheduled to launch in 2017, and has nifty features like a laser ranging system which is ~1000 times more precise than the current microwave ranging system. Because GRACE almost certainly won't last until 2017 (and because climate science requires long-term measurements), the GRACE team has decided to repeatedly turn GRACE off to avoid stressing the battery chemistry. This strategy should stretch out its remaining lifespan at the cost of introducing gaps in the data.
Roderick et al. 2007 shows that pan evaporation rates are dominated by changes in wind speed, with contributions from solar irradiance. You're right to guess that gravity is much weaker than these forces, so changes in gravity don't affect evaporation rates.
You should ask about the magnetic poles and if increasing/decreasing magnetic fields effect gravity at all? Do they take into account the magnetic fields? And if solar storms impact the larger magnetic field outside the planet... Or are they just detecting mass levels?
GRACE just detects mass; solar storms are a source of noise because they exert non-gravitational forces on the satellites which need to be subtracted to analyze the gravitational forces.
My question is if the melting arctic ice cap will impact the mass up there and will cause a change in the gravity?
Since GRACE only detects mass, it can't tell the difference between floating sea ice and the equivalent amount of water displaced by the floating ice. So GRACE can't see the melting Arctic ice cap, but it can see thinning glaciers and ice sheets when the meltwater flows into the ocean because in that case the mass moves.
Does warmer water in the summer weigh more or take up more volume?
Yes, this is called the thermosteric effect; I mentioned annual variations in paragraph 13 of my 2011 paper. Satellite altimeters like JASON and TOPEX/Poseidon can measure sea level rise due to thermosteric effects, but they also measure the sea level rise due to added water from thinning ice sheets and glaciers. Because GRACE can only measure the mass of the added water, it can't measure sea level rise due to steric effects (temperature and salinity). Ironically, this means we can solve for steric sea level rise by subtracting the GRACE estimate of sea level rise from the altimeter estimates.
If snow is less dense than ice (mostly), can you tell where the glaciers are melting and not recovering after the winter?
Not with GRACE alone, but (as with steric sea level rise) satellite altimetry can provide volume measurements. Of course, if the meltwater drains far away from the glacier, GRACE alone could see this change in mass. And, in fact, GRACE is observing this all over the globe.
I've tried to solve for the Earth's nearly-diurnal free wobble, but it yielded a map that looked suspiciously like the mass trend. I think this happens because the NDFW's frequency is similar to that of the K1 tide, which experiences aliasing in the GRACE data as discussed in section 4.4 of my 2011 paper. So... no.
Other satellites noticed Greenland's extensive surface melt because melting snow lowers the ice sheet's albedo. However, water has the same mass as a liquid or solid, so GRACE can't tell the difference between ice and meltwater. GRACE can measure how much meltwater flows into the ocean, because in that case there would be less mass on Greenland.
Also, Ambitwistor referred to the popular monthly GRACE fields, which are available as spherical harmonics and gridded fields. In addition, CNES produces 10 day solutions, and Bonn even produces (constrained) daily solutions. But the monthly fields are by far the most widely used, because the ground track coverage is more complete during a month, and the extra data increases their signal to noise ratios.
As I explained in that link, the Earth's rotational kinetic energy is currently decreasing at ~3.8 TW. (Just to compare, the world used 15 TW in 2008.) But anyone who clicked on that link would learn that only ~3% of the lost ~3.8 TW goes into raising the moon's orbit. As Pete Bender pointed out, the other ~97% is converted to heat in the oceans and the core-mantle boundary layer. Because this percentage isn't necessarily fixed, it's very difficult to predict exactly how much faster the moon would ascend from the Earth.
(Apparently I already answered so many questions that the Slashdot editor doesn't think the remainder will be enough for a follow-up story, so I'll try to answer them here as time permits.)
Excellent links. However, note that the first link doesn't use GRACE to detect Lense-Thirring frame dragging. It's merely using improved gravity models from GRACE to eliminate measurement noise due to imperfections in our model of Earth's static gravity field that could be mistaken as frame dragging.
No, frame dragging is only barely detectable by Gravity Probe B, which was designed specifically for that purpose. GRACE relies on highly accurate timing, which requires correcting for time dilation due to special and general relativity. Here's a reference.
One modest example is extracting energy from the ocean tides. I've explained that harnessing tidal power would actually move the moon farther away from the Earth, even faster than its current ~3.8cm/year recession rate. Tidal amplitudes are influenced by the coastlines and bathymetry, so in principle we might eventually be able to change the tidal amplitudes in some location (bigger for more tidal power, smaller for easier navigation) by carefully modifying the bathymetry.
Just to clarify the summary, GRACE primarily studies long-term changes in water storage. It's just my research in particular that focuses on high frequency signals like ocean tides. Also, the open source code mentioned in the summary is just used to produce my personal results, not the official GRACE solutions.
No, tides are caused by gravity, which is caused by mass. Solar storms are violent, but the amount of mass involved is miniscule compared to that of the Sun. A very powerful solar storm wouldn't even cause a small tidal shift.
Erm, I could point you to someone working on general relativity or quantum gravity. My analysis only uses Newtonian gravity, though special and general relativistic corrections are applied to the GRACE data before they get to me.
Slashdot Mirror
GRACE can resolve nearly uncorrelated mascons that are blocks 400km on each side with a noise floor of ~1cm equivalent water height. (This is latitude dependent because GRACE's denser ground tracks near the poles allow for better resolution.) Each mascon has a mass of ~1.6 gigatons, and a fully-loaded coal train is ~10 kilotons, so GRACE falls short by about five orders of magnitude.
The improved laser ranging on the GRACE follow-on will increase sensitivity, and David Wiese analyzes improvements due to lowering the satellites' altitude and/or adding more satellites to the GRACE system.
You're right to suspect that detecting a tiny change in local gravity is limited by uncertainties in models such as atmosphere dynamics. I've discussed how GPS occultation data (among many other data sources) can be used to reduce these uncertainties.
Other anthropogenic effects such as groundwater depletion can already be detected with GRACE. Rodell et al. 2009 (PDF) and Tiwari et al. 2009 (PDF) observed this in northern India, and Famiglietti et al. 2011 (PDF) recently observed similar groundwater depletion in California.
Also, Barton Paul Levenson explains the World Meteorological Organization (WMO) definition of climate, which uses a more scientifically conservative 30 year minimum timespan.
Correct. Here’s a figure from Royer et al. 2007 which concludes that “a climate sensitivity greater than 1.5C has probably been a robust feature of the Earth’s climate system over the past 420 million years”.
Correct. I've discussed the difference between these situations.
Human ingenuity.
That's a good question; I described the difference between climate and weather at the beginning of my article. I later updated it with a better analogy from NOAA: One way to distinguish between weather and climate is that the climate of your hometown will determine how many sweaters you have in your closet. The weather will determine if you should be wearing a sweater right now.
Many times the climate being discussed is global, so an average is taken over the entire Earth. For global temperatures, Santer et al. 2011 shows that one needs to average over ~17 years of data to obtain statistically significant climate trends. Here's another explanation by Tamino. Also, the Skeptical Science trend calculator helps visualize statistical significance.
That's what I get for writing Australian floods and then editing while dumb.
Yes, it was totally out of left field. I responded in an internal JPL email, and copied the email to my (other) website.
-Bryan Killett, aka khayman80, aka Dumb Scientist
(This was copied from very far down so readers don't have to wade through all these charming comments just to read my answer to eldavojohn's question.)
For those who may not be aware, GPS signals can be used to measure atmospheric properties. As GRACE (or any satellite with a modern GPS receiver) listens to a GPS satellite that's about to pass below the horizon, the GPS signal passes through the atmosphere. Thus the GPS signal is refracted and delayed in ways that can reveal the temperature, pressure, and refractivity of the atmosphere at different altitudes. These are known as GPS occultation measurements.
I've never used GPS occultation measurements, so I asked Gerhard Kruizinga about them at the weekly GRACE meeting. He pointed me to GFZ which has near-real time data (that page shows GPS occultation measurements taken today, but getting the data from GFZ probably requires filling out a quick form). Gerhard also mentioned that GPS occultation measurements from GRACE are regularly fed into the ECMWF atmosphere model, which is briefly described here.
GRACE also measured the 2005 Amazon drought, regarded as the worst in over a century. Just five years later, the 2010 Amazon drought might have been even more severe.
GRACE also measured the 2010-2011 floods in Australian and Columbia, which dumped so much water on land that sea level temporarily dropped by ~6mm.
Reword for clarity:
Warmer water doesn't weigh more (except for absurdly small relativistic corrections), but it does take up more volume. This is called the thermosteric effect; ...
There's no requirement to do so. For instance, ask for the code to the Hubble or WMAP telescopes, or most of the other code produced at NASA. You'll probably be disappointed. I personally think that government funded science missions should release their code in some manner, but I don't want to start a license war. Personally, I chose a license rather than releasing my code to the public domain because I was nervous that a corporation might "invent" my code and sue me for using it. If you'd like different terms than the GPLv3 provides, please contact me and I'll oblige for reasonable modifications.
I've previously discussed some of the advantages of GRACE 2, which is the real successor to GRACE and has a vague launch date in the (hopefully early) 2020s. It'd be great if we could find some way to service and refuel these satellites to extend their lifespans, especially if the low-altitude drag-free option is chosen. But for now we're trying to do what we can with less funding.
Sorry to intrude, but I just mentioned you in a comment. (Also wanted to say that I nitpicked earlier solely to clarify the issue for others.)
But now that I know you're a GR physicist, I wonder if you'd like to read Outlive the stars and leave a comment?
GRACE is dying. Its batteries aren't reliable enough to keep the satellites' microwave ranging system running continuously now. The JPL engineering team has worked miracles to give us 10 years of GRACE data, especially since the mission was only supposed to last 5 years. The GRACE follow-on mission is scheduled to launch in 2017, and has nifty features like a laser ranging system which is ~1000 times more precise than the current microwave ranging system. Because GRACE almost certainly won't last until 2017 (and because climate science requires long-term measurements), the GRACE team has decided to repeatedly turn GRACE off to avoid stressing the battery chemistry. This strategy should stretch out its remaining lifespan at the cost of introducing gaps in the data.
Roderick et al. 2007 shows that pan evaporation rates are dominated by changes in wind speed, with contributions from solar irradiance. You're right to guess that gravity is much weaker than these forces, so changes in gravity don't affect evaporation rates.
GRACE just detects mass; solar storms are a source of noise because they exert non-gravitational forces on the satellites which need to be subtracted to analyze the gravitational forces.
Since GRACE only detects mass, it can't tell the difference between floating sea ice and the equivalent amount of water displaced by the floating ice. So GRACE can't see the melting Arctic ice cap, but it can see thinning glaciers and ice sheets when the meltwater flows into the ocean because in that case the mass moves.
Yes, this is called the thermosteric effect; I mentioned annual variations in paragraph 13 of my 2011 paper. Satellite altimeters like JASON and TOPEX/Poseidon can measure sea level rise due to thermosteric effects, but they also measure the sea level rise due to added water from thinning ice sheets and glaciers. Because GRACE can only measure the mass of the added water, it can't measure sea level rise due to steric effects (temperature and salinity). Ironically, this means we can solve for steric sea level rise by subtracting the GRACE estimate of sea level rise from the altimeter estimates.
Not with GRACE alone, but (as with steric sea level rise) satellite altimetry can provide volume measurements. Of course, if the meltwater drains far away from the glacier, GRACE alone could see this change in mass. And, in fact, GRACE is observing this all over the globe.
I've tried to solve for the Earth's nearly-diurnal free wobble, but it yielded a map that looked suspiciously like the mass trend. I think this happens because the NDFW's frequency is similar to that of the K1 tide, which experiences aliasing in the GRACE data as discussed in section 4.4 of my 2011 paper. So... no.
Other satellites noticed Greenland's extensive surface melt because melting snow lowers the ice sheet's albedo. However, water has the same mass as a liquid or solid, so GRACE can't tell the difference between ice and meltwater. GRACE can measure how much meltwater flows into the ocean, because in that case there would be less mass on Greenland.
Also, Ambitwistor referred to the popular monthly GRACE fields, which are available as spherical harmonics and gridded fields. In addition, CNES produces 10 day solutions, and Bonn even produces (constrained) daily solutions. But the monthly fields are by far the most widely used, because the ground track coverage is more complete during a month, and the extra data increases their signal to noise ratios.
As I explained in that link, the Earth's rotational kinetic energy is currently decreasing at ~3.8 TW. (Just to compare, the world used 15 TW in 2008.) But anyone who clicked on that link would learn that only ~3% of the lost ~3.8 TW goes into raising the moon's orbit. As Pete Bender pointed out, the other ~97% is converted to heat in the oceans and the core-mantle boundary layer. Because this percentage isn't necessarily fixed, it's very difficult to predict exactly how much faster the moon would ascend from the Earth.
(Apparently I already answered so many questions that the Slashdot editor doesn't think the remainder will be enough for a follow-up story, so I'll try to answer them here as time permits.)
Excellent links. However, note that the first link doesn't use GRACE to detect Lense-Thirring frame dragging. It's merely using improved gravity models from GRACE to eliminate measurement noise due to imperfections in our model of Earth's static gravity field that could be mistaken as frame dragging.
No, frame dragging is only barely detectable by Gravity Probe B, which was designed specifically for that purpose. GRACE relies on highly accurate timing, which requires correcting for time dilation due to special and general relativity. Here's a reference.
GRACE measures the accelerating mass loss in Greenland and Antarctica, and tracks global water storage.
One modest example is extracting energy from the ocean tides. I've explained that harnessing tidal power would actually move the moon farther away from the Earth, even faster than its current ~3.8cm/year recession rate. Tidal amplitudes are influenced by the coastlines and bathymetry, so in principle we might eventually be able to change the tidal amplitudes in some location (bigger for more tidal power, smaller for easier navigation) by carefully modifying the bathymetry.
Just to clarify the summary, GRACE primarily studies long-term changes in water storage. It's just my research in particular that focuses on high frequency signals like ocean tides. Also, the open source code mentioned in the summary is just used to produce my personal results, not the official GRACE solutions.
-Bryan Killett, aka khayman80, aka Dumb Scientist
No, tides are caused by gravity, which is caused by mass. Solar storms are violent, but the amount of mass involved is miniscule compared to that of the Sun. A very powerful solar storm wouldn't even cause a small tidal shift.
Erm, I could point you to someone working on general relativity or quantum gravity. My analysis only uses Newtonian gravity, though special and general relativistic corrections are applied to the GRACE data before they get to me.
Yes, it was totally out of left field. I responded in an internal JPL email, and copied the email to my (other) website.
-Bryan Killett, aka khayman80, aka Dumb Scientist