 So we'll hold questions till the end, and Isabella? That's good. I changed a little bit of the presentation to kind of based on what I heard today. I saw that Boyle is coming. Oh, it's not my time. And so there's a lot of text I'm going to browse through, but I thought I was going to bring some example of some ways in which I think about using grace to combine with other, you know, data set and try to leverage on the strings to understand the aquifer. It's the course of my computer. There we go. Okay, so I didn't change the title, but I just, I'm just going to, you know, talk a little bit about, you know, grace as, you know, an integrated measure of total water storage. And so, you know, what are we seeing when we compare? But I'm going to just, this is like a little bit provocatory, you know, fine scale variability with grace. What can we do? And then I'm going to look about, you know, a few examples. We try to see how can we characterize the regional drought, you know, and the evolution and impact using multiple satellite data. And then showing some example of some synergistic activity using drought index. And I think that that's not my time, 220. Okay, sorry. And then, you know, some comparison in some places that are remote and, you know, some work that we've been doing comparing, you know, using grace together, you know, to evaluate our system model, you know, and develop an approach to a diagnostic approach. Okay, so this is like a lot of text, but the idea is that we try to look, we have a lot, we have all this remote sensing observation that we can use as a proxy for, you know, water variability in an aquifer. They are characterized, you know, we talk about, you know, that different sensitivity. Great is the total water storage. Soil moisture is the top. When I talk about remote sensing, 0 to 5 centimeter. And then we look at precipitation. And all of those data are characterized by different spatial scale. So can we use them together and how can we use it in a synergistic approach? And so if you look at the... Do we have it? Oops, no, I wanted to go. Okay, it's okay. If you look on the left, so this is some work that we did for Texas. If you look at the circle in the, you know, in the top right figure, that's the grace footprint. It's about 300 kilometer, 350 kilometer radius. And basically grace give you an average of the total water storage changing in that area. And in this case, we are looking in an area, you know, this is like the land cover for muddies. And, you know, we have like, the footprint is mostly shrub, you know. So we just take a footprint, we have like, you know, dominant by some vegetation kind. And what the example on the bottom left shows, you know, this is like the answer E and 2 soil moisture at the original resolution. And each, so the blue line is the grace average, what grace would see. So the average over the footprint, the red line is the time series of soil moisture in the center point and the gray line on the top part are at every point, every grid cell, you know, within the footprint. And if we remove the seasonal variability, what we find, we find that those lines fall on top of each other. And the idea that means that my average is representative of the inter-annual variability at the single plate. And the idea is that when we come to this, and this is also a function of the fact that we have a surface-sub-surface couple, you know, aquifer, you know, then we can use the information of grace representative, what occurs also at the subscale. So on the right is an example. So we try to say, okay, now let's look at all this dataset and this is for that footprint and try to see can we use this information, this different information to characterize the drought, you know, evolution and the sensitivity of the response of the vegetation. So what we did, if you see the top part of the time series, the precipitation from GPCP, and there's, and I'm just going to go fast here, the idea is that you can see that soil moisture follow very well the precipitation and grace also agree relatively well until the 2011 drought which is a very strong drought and if you look at the vegetation also we have like three proxy use it and after 2011 they kind of got the couple and soil moisture respond to precipitation which makes sense and basically grace has a stronger memory, you know than the recovery and so on the bottom what we look at can we kind of identify a characteristic time scale for, you know, before and after for those, you know, for those for the signal for the time variability and we define it as, you know, as the basically the lag time for which the amplitude of the autocorrelation decrees by 1 over 1 over e and so what we find, we find a very strong drought today about the same lag and then after the response is different so grace has a much bigger memory and the idea is like, okay, so can we use this to characterize this location the response of the system next and this is an example so this is an example using, okay so we define a drought index they are very convenient way to compare, you know, things I'm going to go and this is a grace DSI so we compare it with different drought index the idea is that this provides a different piece of information respect to what we would have with the different, you know, with the other drought index because it's sensible to the total water storage and if you see you can tell for the 2011 grace has a different response to the drought the right shows the correlation between grace DSI and PDSI and it's significant everywhere where it's not the lower value correspond to, you know, the when we had a very big heat wave so we have a drought that was driven more by, you know it was affecting more the surface in mice so grace basically doesn't show this big effect, both the vegetation and PDSI move together and that's a comparison also with groundwater storage I'm going to browse quickly we evaluate this in the US so this is like information that we can use on a global scale we are characterized by an uncertainty and allowed to, again it's complemented to the other drought index so it can be used to interpret the signal I'm going to show quickly this is an example of, you know this is an evaluation this is a very difficult area we're looking at high mountain Asia and can we use grace to evaluate model and total water storage and we define a variable to see the total water storage over the glacier region for the western Himalaya and the Himalaya glacier and the idea is that can we, you know, the implication that if we can evaluate this model then we have more confidence in how we can use it for projection and we can see capture very well the inter-annual variability especially in the western Himalaya and only we have some time where it doesn't miss the peak and in the Himalaya is missing part of the trend and this is consistent with the fact that this model doesn't have ice melting and so in the region where we expect to have more of the ice glacier melt we are missing a trend and the implication of this is that it gives an idea of how the model is representative of the water resources and also, you know, give us information about the process they are driving you know, the change in the glacier and I'm just going to skip this slide and I'm going to show you just another last example of, and I'm over time and this is about you know, this is the comparison okay, just look at the left part and you know, the idea is that can we use grace to evaluate the model output and use it in a diagnostic way to improve the output and in this case this is a water balance model we separate, we separate the upper basin from the lower basin because in the upper basin most of the signal is caused by the glaciers and we compare the model when it's forced by two different forces and you can see when it's forced with the mirror clearly the amplitude of the signal is very different from grace and it turns out the mirror is very dry in this region and that is due to this in the lower basin where we have, you know, what we find the model has what is called is a an UGW represent groundwater is a limited bucket so every time the system doesn't provide enough precipitation of water the water gets extracted by this infinite bucket and with mirror because the input is too dry so we don't provide enough precipitation you can see that you have a very steep training groundwater withdrawal and this is because the model you know, allow the system to extract the water and with the era in time we have a better agreement still we don't match all the trend and this is due, you know we interact with people of the model is due to the fact that, you know we are missing part of the recharge effect of the deep groundwater recharge and I'm just going to skip the conclusion sorry I went long so those were examples of, they could show how you can synergistically use it