 So now we go on to the next part of the program, we have two keynote talks and there's two keynote overview talks. The first one is a talk by Matt Rodel. Matt's the chief of the hydrological sciences branch at NASA Goddard and you know he's there's 65 scientists at the hydrological sciences branch and he got his PhD at UT Austin. He has worked primarily on understanding the terrestrial water cycle, has numerous publications and he's best known for applying the GRACE satellite gravity recovery and climate experiment all over the world in understanding the ground water dynamics. So without any delay Matt and Matt will give about a 33 minute talk and please 33 and a half is not allowed so I'm sorry. And and after that we have about 12 minutes for question and answers and to change over to Holly's talk. So please save your questions till the end and ask a lot of questions to Matt. Matt. All right thanks everyone. I appreciate the opportunity to talk with you all and I'm glad this finally happened after a few speed bumps in the beginning there. I listed myself as the author here. There are of course many many people who contributed to this talk and I hope I've mainly provided you know naming those people when when necessary. So I'm talking about state-of-the-art and future prospects for observing ground water from space. Start with a little bit of motivation here. So global ground water dependence and depletion. So as most of you know and as Tony alluded to there are many parts of the world where people rely on ground water in some cases almost exclusively as their source of fresh water. There have been a lot of papers recently research studies that have shown that ground water is being depleted in many parts of the world. I'm highlighting one here by Yoshihide Wada and colleagues where they used they did a modeling study with with observations incorporated into the model to look at areas of the world shown in the lower left there where the ground water is being depleted at a significant rate. Their model also produces things like ground water recharge at the top middle there and and ground water abstractions. And so when you do the do the math on those it's how you come up with the estimates of the depletion and on the right is showing trends in global water demand which is the top line there. I'm sorry the the numbers are a little hard to read there but the number it's on the left are cubic kilometers per year and the scale goes from zero to a thousand. And then the middle line is is a global ground water abstraction and and then when you do the math on that you end up with the primary depletion which is the bottom line as you can see it's increasing over time the scale there goes from what is it 1960 through a little past 2000 so you can see that that the demand is is has been increasing the depletion has been decreasing and these are big numbers by the way so a cubic kilometer of water is is it's a lot of water so you think of the largest reservoir in the United States which is Lake Mead it holds about somewhere between 30 and 35 cubic kilometers of water so when you're talking of hundreds of cubic kilometers of water it's it's a lot. Another recent study this is one by Dahlan et al looking at one of the major causes of probably the major cause of groundwater depletion which is agriculture and so the left is showing the background map sort of shows aquifers where where there's significant groundwater stress and then the the pie charts show which which specific crops are are responsible or the major crops that are being grown that they're causing that stress and the depletion is is is represented by the size of the that circle so as you can see some of the some of the the drier parts of the world of course there's a larger reliance on groundwater and of course greater depletion and then on the right it's it's a little hard to read up here probably but but this is showing the the the water or groundwater embedded in the international food trade and so you can see for example a lot of food coming out of the U.S. is responsible for for groundwater depletion in the U.S. so let me give you an overview of large-scale groundwater science so in the 20th century not that long ago groundwater science is almost completely reliant on in-situ observations so you mainly using monitoring wells as well as piezometers and doing pumping tests you can use the well drilling logs to look at the the type of aquifer material ground penetrating radar in some cases can be useful and there are other geophysical methods and geological mapping techniques and then in terms of modeling you know there's some some older groundwater models that were fairly crude or highly localized you couldn't really run them over large scales and of course you're limited by the availability of information to parameterize those models. You can do some useful science still though this is a paper by Biling Lee at all looking at data from mainly from the USGS and some others and also from some state agencies groundwater data that were archived and publicly available and you can see it's you know some networks in the eastern U.S. particular that look fairly dense when you're looking at this sort of large scale you know if you got into the smaller scale you'd find that there are plenty of townships where there's not a groundwater observation for example but you can do some useful things like looking at groundwater variability and it wasn't that long ago that people some people even on the planning committee who thought that groundwater variability was basically negligible but if you look at the time series here on the right focus in the middle the black line is the is the mean and this is looking at the upper Mississippi River basin which is shown in yellow on the left black lines the mean and the the orange lines are all the individual wells in that region and you can see that for example in 1988 there was a big drought in the upper Mississippi you can see that very you know stands out very clearly in the well data and then a few years later in 1983 there was flooding and see groundwater gets pretty high it's a peak around then and so it's you know it's you can do some interesting science with these data but it's you know they're you're limited and there's there are only so many wells that that for which data are available so so for example there's you know the USGS archives a lot more wells than I'm showing in this in this figure but in order to do this study we had to we had to sort of do a culling of the information picked out wells are not directly affected by pumping or injections we wanted wells that were tapping unconfined or semi-confined aquifers the reason for that is because it's very difficult to convert a measurement of head in a confined aquifer into a into a change in water storage we want at least four depth of water measurements per year in order to get you know a seasonal cycle and a minimum of a 10-year record when you do that you're cutting out the vast majority of wells that are available for study of course you know we have it fairly good in the US the lower right is showing the USGS groundwater climate response network is which is sort of their network where they've already done this they've boiled down the wells to the ones that are that are useful for climate monitoring and it looks again pretty dense in the in the northeastern US you look at some states like Wyoming and maybe there's only one or two wells there and that in a huge state like that but then you know again we're doing well compared to the rest of the world the upper right is showing that there are only eight countries that that contribute groundwater data to the groundwater global groundwater monitoring network which is at an institution called IGRAC in the Netherlands they collect data similarly to how the global runoff data center is showing the lower lower left there collects data for for runoff and so most countries well you know most countries have some sort of groundwater data but they don't make them publicly available is the issue so things like their their coverage gaps spatially and temporally their delays and and the availability of the data measurement can continue in consistency and but the big things though is is that you know their political restrictions so even if they have groundwater data and even if they are digitized which you know is not always the case and even if they are centralized you know whether or not the general population is allowed to access them oftentimes they're not so moving on to the 21st century we're doing a little better now we have remote sensing particularly satellite revimetry which I'll talk about quite a bit in a moment you also do things like airborne and satellite interferometric synthetic aperture radar INSAR and we have better modeling capabilities driven in large part by you know advances in and computer power and the amount of data that you can store so they're regional and global land service models with groundwater budgets and some of the models are more advanced and have 3d flow capability and then we can do things like coupling those to the atmosphere and and data assembly and assimilation where we constrain the model using the available observations so so let me talk a bit about remote sensing NASA has a fleet of earth observing satellites and and quite a few of them are actually highly relevant to the water cycle and so if you're you have an observation that monitors the water cycle it's probably gonna get some you know in some way relevant to to groundwater studies but some of them are more relevant than others particularly grace and grace follow-on grace mission ended in 2017 grace follow-on launched in 2018 and and those are very those are quite valuable for groundwater this is just an example of what you can do with some of these remote sensed data here the the yellows and oranges that are our precipitation measured by the gpm global precipitation measurement mission and the the blues are showing soil moisture wetness from the smoss mission and so you can see how when when a rainstorm goes over Australia it sort of wets the land surface and then there's this memory where it takes a little while for it to dry out again and when you look at we can only do this really with a model but if you look at groundwater time series you know same sort of animation you would see that the groundwater moves changes even more slowly than this so there's there's a long longer residence time for groundwater and the variability is slow and that's one of the reasons it's such a valuable resources because you can have a drought and still have plenty of groundwater available so let me talk a bit about grace grace is really different from other missions you know most remote sensing missions use some sort of radiation based approach so measuring light is either emitted or reflected from the surface and then that information is used to to to estimate quantities like the snow cover or the vegetation type or ice or rainfall or soil moisture but they're limited in that they can you can only penetrate the surface a few centimeters for the most part so you can't really see groundwater using this type of measurement grace is totally different and grace follow on as well so both of them were were twin satellite missions one satellite following the other around the earth and instead of looking downward at the earth and measuring some sort of you know emitted or reflected or radiation the key measurement is actually the distance between the two satellites and the reason they're measuring that is because earth's gravity field perturbs you know variations of earth gravity field perturbed the orbits of the satellites and so if you can monitor those orbits with extreme precision which is what they do basically one satellite is monitoring the orbit of the other satellite at all times then you can you can basically model how the gravity field changes over time and changes in the gravity field are largely caused by changes in in the water stored on and in the land surface so if you can deal with things if you can use an atmospheric model to to to estimate the atmospheric mass changes their circular mass circulation the atmosphere and you have ocean models to deal with the oceanic mass changes then what's left after you remove those is basically gravitational changes caused by changes in in terrestrial water storage so again the the the key measurement here is the distance between the two satellites so the satellites are you know they they're they're sort of free floating and they're somewhere between a hundred and two hundred kilometers apart most of the time they're measuring the distance with a k-band ranging system there's also a laser ranging system on on grace follow-on using the k-band system they can measure this distance you know this 200 kilometer distance they measure every five seconds down to the precision about the size of a red blood cell so incredibly accurate or precise measurements and again when you put all this data over the course into first of a month into a big regression model you can come up with a map over its gravity field then for month to month see how that how the gravity field is changing in response to changes um particular on in the land surface so what are those changes look like um we call this uh terrestrial water storage which is the sum of all the components of the water stored on in the land surface um if you look at the state of Illinois which is one of the few places in the world where you have ground-based observations of groundwater, soil moisture, snow, and surface water you can put together a time series averaged over Illinois which we did several years ago shown in the top here um where it's it's water storage is an equivalent height of water and what we've done is the blue is the groundwater and then uh superpose on top of that is red which is soil moisture and you see large changes in soil moisture and then white is snow and for Illinois the snow the snow water storage changes are actually pretty small relative to the groundwater and the soil moisture and uh and surface water things like reservoir storage changes average over Illinois are actually negligible you can't even see them um but what you can't see again is that the groundwater storage changes are pretty significant they're not quite as large as the soil moisture on a seasonal basis but on an interannual basis they they can actually be bigger so total trestle water storage is sort of the top line you know the the top contour of that graph and that's what grace is measuring and so if you look in the lower right that's showing a time series of what we call trestle water storage anomalies so it's the difference from the long-term mean at each location so if it's blue that means it's wetter than than the long-term mean if it's red it means it's drier than the long-term mean and so uh again a large part of these changes are changes in in groundwater or large part of these anomalies I should say um one of the limitations of grace though is it can't tell you the total amount of water in an aquifer for example I get that question all the time you know how much water is left in in an aquifer can't can't tell you that can only tell you how it's changing uh over time one of the cool things we can do with um with the grace data look at um long-term trends so if you have a time series uh from grace this case we use time series from 2002 to 2016 we removed the seasonal cycle and then we fit a trend at each location and this shows you where trestle water storage has been changing um on average uh over the course of 2002 to 2016 uh this is in units of uh of centimeters per year equivalent height of water um and then you know sort of a challenge for hydrologists is to determine uh which these trends are caused by the sort of natural variability and they're likely to bounce back at some point uh which are caused by water management or mismanagement and uh which may be associated with climate change so natural inter variability if you want to study that you probably want to look at um how precipitation was changing during the same time period so uh the top right is showing the percentage of normal precipitation um during 2002 to 2016 there are a few places that pop out they got more or less uh precipitation um so areas of increased precipitation i'm circling here on the right uh the lower lower right panel showing um precipitation trends as well and some of these line up with areas uh where there's um an increased in trestle water storage so we might say that you know it was the precipitation change that caused the change in trestle water storage similarly there are areas where either uh precipitation was below normal during this time period where there's a downward trend of precipitation and we happen to see uh similar trends in trestle water storage so again part of the part of the answer to why the water storage was changing um may have been the the changes of precipitation it could be just natural variability and it could come back at some point um if we're interested in you know the climate change impacts you might want to look at the the projected um precipitation changes um that you see in the ipcc reports um so this is showing the uh predictive increase of precipitation increase or decrease in the um the high scenario uh you know a lot of carbon dioxide at the end this is by the end of the century um and there are areas where uh where we see an increase a predicted increase in precipitation there also happens to be somewhat of an increase in in um trestle water storage so perhaps this is a sign that there's a climate signal in the trestle water storage data and similarly areas where there's a predicted decrease in precipitation and we're seeing decrease in trestle water storage and then finally uh the water management um irrigation is the is the largest user by far of trestle water storage and groundwater irrigated crops and there's some areas of the world where we know that irrigation is extremely um intense and those um some of these areas I've circled really line up very well with where we've seen trestle water storage depletion and there have been quite a few studies um that have focused on these areas showing that yes indeed the groundwater is declining one area that's interesting though is there's an increase here I've I've uh used an error to point to increase in the trestle water storage data which is an area of China where the Three Gorges Dam and several other dams have been filled over the past two decades and uh and you can see an increase in trestle water storage that's caused by that um one of the first areas we studied when we're um looking at um the trends in trestle water storage um Isabella Vellaconia and JFMA I looked at uh Northern India back in 2009 um we've done we've looked at it again more recently and there's still a significant depletion of trestle water storage in that region uh there's also there are well data and other you know other reports information where we know um with with high confidence that this is reflecting groundwater storage depletion so much water is being pumped out of the aquifer to to irrigate crops a lot of those crops are are water-intensive things like rice or wheat and uh it's causing the the groundwater to decline rapidly we estimate rate in this this area of circled about 19 cubic kilometers per year so so every two years or every uh every three years are using two lake meads worth of water in this region and it's gone you know they use the water the crops it's and it most of it evaporates some of it runs off but it's not this is the water that's not recharging the aquifer it's it's gone to the ocean and river else um in India you know we when we published that study it was um it you know it was it was shocking and and well received um but it wasn't like there wasn't data available already um more people could have figured this out they it's hard to even see here but there's so many little tiny dots here that are that are um uh they're dug wells or piezometers um all over India but you know the data were not made available to the public so people couldn't do this kind of study without using a remote sensing data um back to the emerging trends uh we published a paper last year um that looked at uh the causes of these trends in terrestrial water storage around the world um you can find this paper in nature last year if you want i'm not going to go through all these um but suffice to say that a lot of these um depletion trends that we found are caused either in part or wholly by groundwater depletion um so talking about sound like ribometry let me just talk about the you know it's it's not a panacea right so there's issues with it one of the big things is that the spatial resolution is very low so talking about areas um larger than about a hundred thousand square kilometers you know at best um size of the state of Illinois is about 150,000 square kilometers so we can barely get you know look at an area smaller than the state of Illinois the primary limitation there is um what we call spatial temporal aliasing which basically means we're not measuring each location on earth with grace or grace follow on often enough and there's high frequency variations in in the atmosphere in the ocean that then alias into the these monthly averages that we get and we really need um more observations which would mean more pairs of satellites in order to address that uh monthly uh monthly temporal resolution is an issue in some cases uh data latency if you're going to use grace data for an operational uh application um that can be a problem um the standard data latency for grace was about two to four months um with grace follow on they promised to have a what they call quick lift product which should be available within about two weeks and there's also this lack of vertical information so I showed you before that tracer water source the sum of all the components of of water sort on and in the land surface and and grace gives you no information on whether the the change in the tracer water source was in the groundwater or the soil moisture the snow or or whatever um opportunities uh Tony might be interested in this so um again if you want to address the the low spatial resolution you'd really need to have multiple pairs of grace like grace like satellites um there are new technologies that are being studied one is um using laser interferometry which is what grace follow on has but if you pair that with uh lower altitude satellite um which would have to be drag free meaning you have a drag free propulsion system on it um then you can get potentially get some higher resolution although that doesn't get past this uh aliasing issue another new technology is being studied something called cold atom radiometry where you have a single um satellite system and you're actually measuring how the gravity affects um atoms as as they're uh as they're moving around within the satellite and this is uh um this is something that's being uh technology is being developed like by Goddard and others um maybe there are other news technologies I'm not aware of um and finally data simulation is a tool that we we're already using to um to address the spatial resolution latency etc. I want to talk about one more um remote sensing technology which is INSAR I mentioned this before so INSAR relies on a satellite um making a measurement of the of the uh the elevation of the surface and coming back again looking at how the elevation has changed um and uh and the reason this is important is because when when the groundwater level changes in an aquifer um it's sort of holding apart the the aquifer media that the you know the pores are being pushed on by the water and so when you remove water there can be compaction and so over time if you're depleting an aquifer you're going to see the the the land surface subside over time it's a famous picture on the right um actually from 1977 showing where in the San Joaquin Valley in California where the land surface used to be before they're pumping all this water and you know by uh 2019 he'd probably be this guy'd probably be you know 30 feet above the surface um and uh so again what's happening is the aquifer is compacting this is what you're measuring with uh with the INSAR is the is how the the land surface is changing and we may be able to do something in terms of um estimating how much water has been removed now the issue again is that you know you take water out of the aquifer when you put the water back in it doesn't it's not elastic so it doesn't come back up as much this this graph in the in the lower right shows you how it sort of changes over town there there's more there's more down than up um so it's it's difficult to uh to directly link a change in the land surface height to a change in in the aquifer water storage um so here's some some imagery you know the the big advantage of this of course is you get extremely high spatial resolution so so here's uh from Palsar uh here's uh changes in the um in the height of the land surface from july 2007 to december 2010 and then may 14 to january 2015 these are turning times of drought and then may 2015 to may 2017 he's in pretty huge changes and the scale over there goes up to 60 to 70 centimeters of of subsidence and this is showing some time series for some of these regions of how the land is subsided and then uh there's some rebound when you do get finally get some some rain recharging the aquifer but look at the scale over there it only goes up to four to five centimeters of rebound um and this is uh credit tom far at uh nasa jpl who developed these maps so in sar you know again i i think i've already covered this but um the big advantage is the spatial resolution but the main issue is that um you have this non-elastic aquifer response there are also issues of things like when the vegetation grows or you plow the field you're changing the heights a little bit and then that's you have to try to interpret that so um future freshwater remote sensing um i'll just mention that the uh national economy's decadal survey that came out last year um and some of the key observables that were recommended well there's the the surface water ocean topography mission or SWAT which will measure um surface water and and hopefully be used to estimate the rates of river discharge uh that one's already developed and the decadal survey recommended uh another precipitation measurement precipitation measurement mission mission which is um which is critical if you're doing any kind of hydrology i also recommended another mass change mission so follow on to grace follow on so that's great news for for us who are interested in groundwater and then further down in in their uh in their report where things like snow depth and snow water equivalent planetary boundary layer um so i'm not sure about transpiration um another thing to think about is that we shouldn't necessarily only be uh considering these these large sort of flagship space agency missions um there are a lot of other types of observations that are valuable for monitoring the water cycles they can put sensors on commercial aircraft for example um already we're putting sensors on the International Space Station um they're all sort of ground-based techniques you can use citizen science you can uh you can use information on the how fast the the signal goes from your cell phone to a tower to estimate uh precipitation rates so there are a lot of other things that that we can be doing besides just looking at uh flagship missions and this is summarizing this um paper by Matt McCabe at all that came out in 2017 so let me talk a little bit about modeling and data simulation a land surface model is basically a sort of like the the land component of a climate or weather forecast model divides the earth up into a grid and at each grid point it may subdivide it into different vegetation types and then you have equations that represent all the processes you know what happens to the water and the sunlight after it hits the after it hits the land surface and so um we can do things like uh assimilate the grace data into into a land surface model so the land surface model provides the high spatial and temporal resolution um provided by the the model parameters and the other inputs things like precipitation and solar radiation and then use the gray state to constrain the land surface model and so when you combine the two you come up with something that's better than than either of them so that the top panel the animation here is showing that grace tracer water storage anomalies and the lower panel is showing um what happened you know the model output when you assimilate those anomalies and you can see there's a there's a higher spatial resolution in the model output and but overall the patterns are generally the same um there are more advanced ground or flow models that are available now uh mentioned a couple here one is one's called power flow um laura condens here and she was one of the key developers of power flow along with reed maxwell at um at um csm and uh and then the right shows the pcr globe wb model um which in the very beginning of my talk i showed some results uh from uh yoshi adwada that were based on uh based on this model um and uh and both of these models have the advantage over what i was showing before um in that they can they can simulate 3d groundwater flow so not just not just the changes in storage of some of the more simple land surface models do but actually how the water flows um up and down and in all different lateral directions um some of the various observations that are useful for groundwater modeling um you know there are a lot uh the the precipitation i mentioned is key but any of these other things like soil moisture or stream flow they all provide constraints that we can then use in our analyses or assimilate into our models to improve our estimates of uh of uh groundwater storage um and then our wish list would be things like um the root something you know sort of deeper soil moisture um higher resolution choice of water storage etc so here's my summary of 15 seconds left so so um i'll let you read this um maybe i can talk about a little bit here that so groundwater is a vital resource as you know that's why you're all here um it's hard to find long-term reliable in-situ groundwater data and that's why um remote sensing and modeling are are important um advances in remote sensing could eventually improve spatial resolution accuracy and timeliness of what we currently have from from things like grace um and but in the meantime integrating the data into a land surface model um is one way to downscale and disaggregate and interpret these water storage observations that we get from from uh from grace grace follow-on and insar etc so thank you thank you so much for staying on wonderful time you have a few minutes for questions i know that was a lot let me keep talking for another 12 minutes here no no hi matt from the us map of groundwater well observations seemed like some of the states are barely being monitored at the same time that map from india showed that there are quite a few wells i mean quite a few observation wells all over the country but that data may not be available to us scientists is that where the problem is that even if the world is i mean even if groundwater is being monitored all over the world there's no no consensus of bringing that data together is that where the problem is there there yeah ali akanda i'm from the university of flood island i'm an assistant professor working on water resources and groundwater is one of the areas that i focus on so so you're you're correct um there are multiple issues here now if i show the usgs map of all of their wells or all you know all the wells that all the states archive it would be much denser but we've already we've already just removed a lot of the wells that don't have a long enough time series some of them maybe just have one observation ever and we've also removed the wells that are in confined aquifers or wells that are affected you know directly by by withdrawals so when you do all that you end up with a much smaller number the map i showed of india i hadn't done they yet so presumably if you're if you if you narrow that down to just the really useful wells there would be a lot fewer wells but the issue remains that that most countries of the world while the governments may have access to data um they don't make them available but before you even get to that point i mean maybe the data is all you know all on paper no one's taking the time to digitize them yet um or maybe if they have they're you know they're they're all over the place are not centralized in one location and then finally do they make them available to the public in some cases like like in india it's become much easier for um indian citizens to to get access to the data but but i still can't get it myself as a u.s citizen so so a lot of issues contribute to the problem and of course there are areas like you know like iran which may have significant groundwater issues but good luck getting any groundwater data from them right uh ed bigley northeastern university could you say a little bit more about um you were talking about the resolution of the grace data with the right now we have one pair of satellites up there you know if we had two pairs like what or three pairs or four what kind of spatial resolution could we eventually get down to yeah so if you had a you know two or three pairs um i think you would still probably only be getting down to you know maybe half of where we are now so maybe be at 50 000 square kilometers set of 100 it that's that's again it's the biggest it's the biggest issue in terms of um spatial resolution accuracy is is a spatial temporal under sampling but to get down to really you know much higher resolutions we would need some of these more advanced technologies that i that i brought up isabel do you agree with what i said there i i forget the numbers on if you have two or three pairs yeah i think with two pairs there you would already get yeah you get you could get down probably at you know a scale which is like halfway with no 50 kilometer but how many but i think with two pairs and one in the almost polar orbit and one with a lower yeah orbit but the issue would be the the fact that we have to remove all the some monthly signal however it is true that eventually those model you know once we have the data as soon as we get improved and that you know we can reprocess data and get information so that would really i think be a big game changer i also think that there are some approaches when you use the grace data that you can you know leverage on you know again try to see you know if you can the resolution is there but maybe because of the spatial variability of the signal the time variability on the signal that you want to matter you can isolate it even in a big for pre i mean as long as the signal is cross correlated you know at those scale then you can so i think that there is a lot of potential also in try to identify you know how can we push the boundary using this isabel a velikonia from yours of california if you don't use the mic and you won't be recorded all right thank you so tony laid out several issues logistical issues for the armed forces trans boundary water rights and so forth just you know at one level i think it is breathtaking to see a global view of groundwater or total water water storage on the other hand if i wish to inform the questions that tony asked i'm struggling a little bit uh you know i can see these are proxies but i don't see how exactly they work uh but maybe there is a future and if you could talk a bit about how these map over to those sort of questions yeah so i mean there are definitely things we cannot address for example tony was mentioning you know where is the groundwater you know sorry can't really help you there um you know but things you know for sort of a policy discussion on you know how is groundwater changing and how might that affect um you know regional conflict or or movement of people and that sort of thing um i think we can we can provide some information on how the groundwater is changing over time and in particular as i said if we use the models or we use these more advanced technologies like uh like insar we can get down to the resolutions um that are that are useful you know i agree that you know the global map you know i don't know how much help that provides to tony but but when we start doing the downscaling using the uh using the lancifers modeling and if we applied insar we had one of these advanced technologies um i think we get to the point where we have uh information that is valuable for some of the things that that tony's asking about as such as our from the university of Oregon um i work on insar groundwater monitoring um i really like your your part of data and the integration because we we are now having monitoring with grays which looks at gravity inside that looks at the formation we have smat and swat so my question is like do you think we have the capability in modeling to integrate all these different data sets together to really take advantage of the multitude um data we have from remote sensing or do you think we're not there in term of the modeling capability yet i think we're we're getting close so for example the land information system which is um a modeling system developed at at nasa goddard um it's basically a framework they freaking run multiple different um lancifers models and some of them are better for for modeling groundwater than others and it also allows you to do data simulation we can already assimilate smap and smos grace uh leaf area index um snow cover and and uh snow water equivalent uh we haven't gotten to the point yet where we can do we can assimilate insar information um and i think that's going to take take a while we're going to get better you know at insar and then also in our models we have to assimilate we'd have to be able to simulate um how the the surface variations change and how that relates to the to the groundwater storage but to answer your question i think we're we're getting close we can do a lot of it we can't do all of it yet but uh we're continuing to push and we actually have a um a new uh proposal where we're um proposing a couple par flow which is one of the more advanced models i showed um with the land information system so you'd be able to run you know using uh you know a very sophisticated uh groundwater model and then also do the data assimilation um uh and other techniques that we we have available with lis so last question jim dobervolsky i'm a national program leader for uh usda national institute of food and agriculture i hear in washington dc for the moment um what i'd like to know is what about basic water quality differences like fresh water sea water uh so that we could actually evaluate maybe how um the use of of you know water in managed aquifer recharge might help us to push back saltwater wedges or what have you any any help there not much i basically nothing that i showed really takes water quality you know calendar has the ability so the remote sensing you can't do anything with water quality it's just you know massive water or or a change in surface elevation uh the modeling i think you know maybe at some point in the near future we get to the point where we're incorporating water quality into some of these these groundwater models but we're not there yet as far as i'm aware um but it is you know it's an incredibly important topic i i just uh maybe there are other people in the room who can provide more help than i can thank you very much thank you matt thank you thanks