 Hello, everyone. I would like to welcome you all to our sixth and final webinar in our Water Productivity Masterclass series. My name is Lauren Zalinski. I am part of the Water PIP team from IHE DILT and I'm joined by Abraham Abishak, also part of the project team from Mehta Mehta, and we will be moderating this webinar today. We're very excited to see everyone here, and if you could introduce yourselves in the chat box, we would really appreciate it, so you can put your name, the organization you're working with, and the country that you're from, and it helps us understand who our audience is today. So, if this is your first webinar, I would like to tell you that the webinar series is brought to you by the Water PIP project, which is funded by the Dutch Ministry of Foreign Affairs, and it brings together different organizations that focus on water science and water management for the purpose of improving water productivity in the agricultural sector. So, most of the activities are being carried out by a group of three organizations. They are IHE DELT, Institute for Water Education, Bacheningen University in Research, and Mehta Mehta Communications and Research. So, we're very happy to have you all join us today, and if you are returning attendee, welcome back. It's nice to see you. Like I said, this is our sixth and final week, so it's been quite a journey. We started in June with an introduction to water productivity and how to monitor water productivity. The next two weeks, we focused on using the Wapour portal and the data within that portal to monitor water productivity, and then we switched to talking about a specific crop, talking about sugarcane production and concerns around water productivity. Last week, we focused on the social and economic parameters around water productivity, and finally this week, we will be learning about aquacrop and how we can use that software to also monitor different parameters around water productivity. If you would like to re-watch the recording from today or download the presentations or do so for any of the previous webinars, you can go to the project website at waterpip.un-ihe.org, or you can go to the waterchannel.tv. So, both of those websites will have the videos, and then the project website, you can also download the presentations and find additional information. The project is also on social media, so if you are on Twitter and would like to follow us, our handle is at waterpipproject. There we put out notifications on new events and new publications or project reports that we put out as part of the project, and we're also on YouTube, so you can subscribe to our channel on Waterpip Project, and there you can find all the videos from the webinar series, and in the future as we have more events, you can find the videos there as well. So, the agenda for today will we will have presentations from our colleagues at Vakeningen University in Research, so first we will have Maria Cristoforado, and she will talk about the agronomic aspects of water productivity and then followed by a simulation of the aquacrop software, and after that she will introduce a case study that we've been working on in Kenya, and then that will be followed by Hararovon Palsama, Associate Professor at Vakeningen, and he will talk about a diagnostic analysis of that case study in Kenya, and then also talk about a comparison between aquacrop and the WAPOR results as part of this this effort, and as always after the presentations we will have a Q&A session that will be moderated by myself and Abraham, so we will not pause in between the presentations, but if you have a comment or a question we encourage you to put it into the chat box, and while the presentations are happening Abraham and I will be collecting the questions, and then during the Q&A session we'll put them on the screen and present them to our panel for a discussion. So with that I think we're ready for Maria to begin her presentation, and thank you all for joining us. Thank you Lauren, thank you for your introduction of today's webinar. So I'm Maria, I'm working in the waterproof project as a research assistant with Vakeningen University, and today I'm going to talk about how aquacrop software works and how it can be used to do a diagnostic analysis and understand the reasons behind lower high values of water productivity. So at the beginning firstly I'm going to talk about these agronomic aspects that influence water productivity, such as feeding management, water stresses, and other forms of stress, then I'm going to present to you how aquacrop works and the simulation steps that it uses, and then I'm going to introduce the Cropman case study and the methodology that we used, and then Geraldo is going to discuss more in detail the diagnostic of each form that we simulated and present the aquacrop vapour comparison of the results. So we have seen in the previous webinars that water productivity is about the kilograms of biomass and kilograms of yield over the total amount of water that is evapotranspired, the water that is consumed. However, in turn this biomass in the yield is depends on the physiological processes of crop growth that affect transpiration biomass and the determination of the harvest index. The harvest index is defined as the percentage of the harvestable biomass against the total biomass production which of course defines the obtained yield that we have in the field. These aspects, the transpiration, the biomass and the harvesting index are also dependent on agronomic field factors such as water stress, temperature stress, field management practices, soil fertility and soil salinity stress. For example, when there is severe water stress, transpiration cannot take place and thus biomass is not produced. In turn, depending on the timing and the duration of this water stress, the harvest index can be either positively or negatively affected and affects also the yield. Additionally, we know that the initial soil water content is very important for the germination of the seed and possible shortages of water during these first days of crop growth can cause the complete die of the plants. Lastly, choosing a correct variety and appropriate variety depending on the climate is very important because it is not logical to use a variety with long growth cycle under rain-fed conditions and dry climates. Overall, all these agronomic factors affect the physiological processes of transpiration, biomass and the determination of harvest index and thus also affect water productivity. For this reason, in order to understand the reasons behind low or high water productivity values and direct our interventions in meaningful ways, we should understand and analyze the influence that these agronomic factors have. So Aquacrop provides an easy-to-use software that considers all these factors that I mentioned before. Aquacrop was developed by FAO and is a plant growth model that simulates transpiration, biomass and yield with a relatively limited number of inputs required. It gives a day-by-day water balance in a sort of vertical transect that studies the interactions between the soil, the plant and the atmosphere, the climate, as you can see in this graph. Aquacrop also separates the evaporation from the transpiration part of ET and thus it calculates biomass based on transpiration and does the beneficially used part of evapotranspiration. Aquacrop can be used in herbaceous crops such as barley, cotton, maize or wheat and is freely available in the link that I have put in the bottom of this slide. So let's have a look at the steps that Aquacrop follows in order to do the simulations. Firstly, the green canopy cover is simulated as a percentage of the soil surface that is covered by the canopy. In general canopy development is characterized by four growth stages, the emergence, the maximum canopy cover, the senescence and the maturity. The green canopy development also depends on two different aspects. The first one concerns the physiological properties of specific crops such as the type of the crop and the variety as well as the climatic and management conditions which involve the stresses, the water stress, the temperature stress, soil salinity and soil fertility stress. Based on these aspects Aquacrop simulates the actual canopy cover development as well as the optimal green canopy cover development that would have taken place in the absence of water stresses or other different stresses. In this sense Aquacrop can also provide insights in terms of production and productivity gaps. In the second step the transpiration is simulated. As we see in this brief flow diagram canopy cover is also used in order to simulate transpiration. Transpiration is also affected by severe water stress as I mentioned before severe water stress or aeration stress due to excess water can cause stomatal closure and thus reduce transpiration. Additionally transpiration is affected by soil salinity and cold stress. Moving to the third step that Aquacrop uses is the simulation of the biomass. As I said before Aquacrop calculates biomass production based on transpiration and thus the productive portion of evapotranspiration. Biomass is proportional to the cumulative amount of water transpired so the more transpiration the more biomass production and thus there is a linear relationship between those two aspects. Aquacrop uses this this aspect to calculate the biomass water productivity star which is normalized for the climate and the CO2 concentration. As we see in this in this graph the biomass water productivity star has a different range of values between C3 and C4 crops so Aquacrop also distinguishes between the photosynthetic capacities of different types of crops. Safer T stress affects also this photosynthetic capacity of the crop and thus the biomass production in total. This can also be seen in the equation that is just below the the graph as soil fertility refers to this KC factor. The fourth step of Aquacrop simulation is simulating the yield which is based on the harvest index. So the harvest index depends on two can can be changed based on two different parameters let's say. The first one refers to adjustments that take place due to cultivar specific requirements and are taking place for example when we have a failure of pollination or insufficient green canopy cover that results in that the crop cannot reach the maturity level. The second set refers to water stress severe or mild that can have either positive or negative effect on the harvest index. As we can see in the picture stresses before the productive phase and mild stresses during the yield formation can increase harvest index while more severe water stress during yield formation decrease the harvest index. So having said that having understood a bit the background of how Aquacrop works I will introduce the case study the Cropmon case study that we apply the Aquacrop simulation model. Our analysis was about two commercial wheat farms and three subsistence maize farms in Kenya and our our calculation were based on the field visit and master thesis research of Michiel Koesters of 2019. So in order for Aquacrop to assess the climatic and management stresses we need to have certain input files that we use for the initial simulation. So as a first step we input it a climate file with temperature and rainfall data that were obtained from the field from meteorological stations next to our farms from Tahmo organization. Then regarding the specific crop characteristics we inserted the planting date and the planting plant density based on field data. Regarding the rooting depth we we assumed based on the literature a rooting depth of one meter for wheat farms and 0.9 meter for maize farms. Lastly regarding the specific crop growth stages we used the default Aquacrop values for wheat and maize that defined the time that it takes for the crop to reach the next stage. In the next step we had the management file and since all farms were rainfed we didn't include any irrigation schedule or irrigation inputs. In this in this file we also inserted the soil we assumed some soil fertility stress and for for the wheat farms since they are commercial we assumed a relatively good soil fertility so we assumed a stress of 15 percent soil fertility stress and 10 percent of wheat infestation while in the subsistence farms maize farms we assumed a level of soil fertility of 40 percent and 15 percent of wheat infestation. As I will discuss later we we used this parameter to validate the results of Aquacrop with the field data. Lastly in the soil file we added the soil type based on soil classification maps and since the rooting depth of wheat and maize is relatively small we assumed no interaction with groundwater table. So after the initial run with the previously measured parameters we validate Aquacrop's results based on canopy cover field data of 10 meter resolution and the reported yields from farmers interviews. The parameters that we mostly used in order to validate was the soil fertility and wheat defense infestation and for the initial soil water content we assumed that soil water at the beginning of the planting date was at field capacity and thus it was enough water was in the root zone to to avoid water stresses during this initial period. Regarding the crop growth stages we had to adjust slightly the date of maturity so that we have a better fitting between the observed the reported yields and the simulated yields. And as you will see now following Gerardo is going to discuss more in detail how we did the simulations and the diagnosis for these farms. So I give the floor to Gerardo. Thank you Maria. I will just take you along the farms that we've been simulating in Kenya first to start with the commercial wheat farms. Rain-fed conditions and as I said we have two sets of field data that we can use to validate the simulation of Aquacrop. As you see in here in this graph my pointer works. Does it work? Okay I can't see myself but so here you see these black dots and those are basically the canopy covered data that we obtained through the Neo satellite images over the growing season. So that gives an idea of how did the canopy develop in this field during the growing season. The other set of data that we have to validate is the reported yield by the farmer. So in this case the 3.34 tons. And as Maria explained we have a given climate we have the soil data that we put as an input and then we can play with either the soil fertility or the wheat infestation to try to come up with the best match based on the canopy cover data that we have and the reported yield. And we're not touching yet the internal crop specific parameters that define the growth stages because basically we don't have detailed enough information to do that. But in this case what we see and see the results is if we set a soil fertility stress of 25 percent and a wheat infestation of 10 percent we get a reasonable result. The observed yield or the simulated yield is very close to the observed yield and we are fairly okay it's not perfect it's a reasonable simulation of the canopy cover. So the simulation in our crop seems to work quite nicely. And what it does enable is to look at how does the growing environment perform over the growing season. And over here you see a lot of information basically on the top bar you see how much transpiration has occurred during the season and as long as it is blue it basically says you meet the crop can transpire the amount that is required. And here towards the end of the season we see that the blue bars do not match the potential bar and basically we're facing a condition of quite severe water stress leading to stomata closure and that is we can see immediately here in the graph underneath which is basically an indication of how much water is available in the soil in the root zone and here towards the end of the growing season. We are building all the red line threshold line for stomata closure and we have quite some severe water stress which is indicated above it shows that we have too little transpiration going on. Looking at the canopy cover immediately we see that we stay below from quite from the beginning from the season we stay below the potential and that has to do with two factors that has to do with the fertility stress that we're imposed we're in this case 25-15 percent fertility stress but also during the early stages of the growing season we see that the water level drops below the green threshold line what we call the mild water stress that is basically the crop is losing its turbo pressure it does not have enough force to really expand its canopy development and in aquacrop that's defined as the canopy expansion stress so that is another water stress limitation on the growth of the crop and so basically we're saying quite we see that throughout the season the canopy cover stays substantially below the potential canopy cover and that will have an effect in terms of the biomass that is being produced and the yield that is produced so we have a biomass production only in this case of 7.9 tons per hectare and a yield of 3.45 tons per hectare and we can see that different stresses are playing a role in here so it's the fertility stress that affects the amount of canopy that can be developed and developed we have in the early stage of the season we have actually too much water too much rainfall which is leading to a little bit of a radiation stress then going into the canopy expansion of the emergence we have mild water stress which restricts our canopy development and later on in the growing season just as the flowering season is starting we basically have coming to an end we have a severe water stress which is leading to stomata cloaks so it's a combination of stresses that all accumulate and have their effect in terms of the total biomass that is being produced below so on the harvest and in this case what we see is that the harvest index which was potentially 48 percent is the default value that is that is adjusted to be 43.6 percent which is mainly due to the severe water stress that we see occurring from flowering and later it also provides us with an opportunity so looking at what could you do in such a condition in such a case and basically what are the management options for a farmer in these conditions which are rain-fed most obvious is of course to eliminate the fertility stress so what will happen if we would be able to provide enough fertility to this crop would that help and seemingly yes because according to the aquacraft simulation we will get a higher yield from 3.4 to 4.3 tons per hectare but it is not as much as we would potentially anticipate and it is basically because we still face that water shortage and the shortage in rainfall during the second half of the cropping season and in effect the water stress becomes more severe under no fertility stress than under in the case with fertility stress and that is because just the other way around so here the water stress is severe without fertility stress than with fertility stress and why is that because in this case now with no fertility stress we can immediately see we have a full canopy development in the early stage of the crop season but we are basically consuming more of the water stored in our soil at an earlier stage of our growing season so providing fertilizer in this case has some effect and it will have a measurable effect in increasing the yield but it will not increase it as much as you would hope for simply because we don't have enough water last during the second half of the growing season to to provide a higher production we see similar situations happening in the other farm the wheat farm and I think here we are more happy and exactly we assimilated in the in the same manner so again we have the canopy cover data throughout the growing season that we used and we have a yield that has been reported and in this case we were able to get a much better match between the canopy assimilated and the canopy that we have observed with the satellite images so we are quite confident that the default settings for the wheat crop as they are provided in aqua crop where they define the growing stages in growing degree days is working quite well for this case of the wheat farms in Kenya so basically we are validating that that basic crop file for for wheat in this case and also here you see the similar situation as in the in the previous wheat farm where we have a small limited reduction in the canopy because of our fertility stress but as the season now then we have some mild water stress in the beginning of the emergent stages which reduces the canopy development and later on we have some severe water stress which is really reducing again the the transpiration so different stresses occurring at different stages over the the growing season all having their effect on biomass on transpiration but also on the yield and in this case we have again due to water stress we have a mild effect on reducing the harvest index which is by default set at 48 percent in the aqua crop settings but because of water stress we have a reduction of the harvest index to about 46 percent which is mild the other reduction is basically just that we have not enough biomass produced to provide a higher yield due to a mild fertility stress and a quite pronounced water stress what we like with with the application and what we show with the second farm is that the default crop settings for for the wheat in this case seem to work well we'll see now when we're looking into the maize fields we're coming to a much more difficult situation which we basically try to do the same approach we have a set of canopy covered data which is very long and there is quite a bit of of noise which a section of this canopy data is is green we don't know we have a yield that is being reported and we try to match it and in this case as Maria was already indicated with that okay we the first step we can switch see with how much fertility stress can we use subsistence level small scale farming we know the fertility stresses occur a lot in east african rainforest conditions so we've opted for severe fertility stress in this case up to 40 percent which gives a not a very good match but the statistics are not good the other thing that we're facing is that that we face in this simulation is that we find it was difficult with the default settings of the maize crop to come close to the observed of the reported yield of the farm with the simulated to this end basically what we did needed to do is to change the crop settings within aqua crop and the only way we could come up with a a semblance between the simulated yield and the observed yield is by extending the maturity time for the crop by in this case a hundred growing degree days and basically now we are going to enter into a dangerous situation from a crop simulation point of view because yeah we're manipulating the basic crop settings within aqua crop and normally what you would like to do is to have do that on a proper calibration where you then have detailed information exactly about canopy cover but also on the different growth stages in the field so that you know okay now my crop is starting emergence on this day and the canopy cover is so much and it starts with flowering on this date and the canopy cover is so much and the same for maturity and so on but in the absence of this detailed data where we have a very clear link between the canopy cover the date and what's growing stage the crop is basically we lack the information to do a proper calibration of the canopy development for this crop so we did a slight manipulation it's okay if we increase the maturity by 100 growing degree days but basically we don't have the good field data to say that then we get a reasonable return in the yield yeah but the simulation in terms of canopy cover is is not so good but okay I guess an idea what may have happened and I think here you clearly see that the contrasting situation is we have much more water during the growing season and actually we have extended periods like here and here where we have too much water in the field which reduces transpiration and with that also some of the biomass production because of too much water in in the soil the canopy is is reduced primarily because of this severe fertility stress that we have introduced another issue that we face in this context is when we look at the yield and the harvest index of the situation there is a discrepancy between when the crop is mature in the simulation where you say okay after 154 days after selling the crop is matured and basically ready to be harvest however in the field the farmer in the reported harvest was 243 days so that's a substantial longer period now also from literature and from from other studies we know that it is not uncommon in some of the conditions of East Africa for farmers to leave their mace crops standing on the field after maturity to dry and to really to dry the crop standing out in in the field but this has risks also with regard you can lose some of your yield due to diseases and rotting if you leave it standing out too long in the field to dry out from literature we we know some farmers can leave it up to 90 days after maturity in the field but from a simulation point of view this complicates matters a little bit because the reported yield is that then does that include yield losses only during the growth stage so was that already yield loss at maturity or is that a yield loss that has occurred after the maturity has as a breach in the lack of detailed information this this complicates matters a little bit in this case we have a slight reduction in the harvest index and as you will see this is not due to water stress because basically we don't we're not facing a lot of water stress in this condition but it is purely because we have extended the maturity day date by growing degree days and then aquacrop simulation indicates it's okay at a certain point I don't have enough green canopy left to produce more yield so potentially I would be able to produce more yield but I just do not have enough green canopy left on my in my crop to produce more more you so we have a slight reduction going from about 48 percent to 50 45 percent a similar situation in in the maze in another maze field again we need to assume very high levels of fertility stress in order to come to some reasonable amount of of a fit between the canopy simulated and observed and the yields in this case however to really come up with with a good fit between the yields we had to extend the maturity even more to with 500 growing degree days and we know that the maze variety that is grown in this farm a is different than the maze variety in farm b whereas for the two wheat farms we know that both farmers used exact same variety of wheat now we're dealing with with situations of different varieties which we would expect to have slightly different settings in in their crop settings in in aquacrop so we don't have a very good simulation of the canopy cover but in the absence of additional detailed information we're not able to come to a proper calibration of the crop settings for for this case this is what we have to to do with so I have to be a little bit cautious with interpreting the numbers and and and the results interesting here I think if you reflected to the other maze farm we clearly see okay we have again we're dealing with the issue of fertility stress that we induce which is reducing the amount of the canopy but we have much more water scare situation to it's the end of the of the season with poor rainfalls occurring so that is affecting the the biomass and may also affect the growth stages and in addition to that we have temperature stresses where there is considerable cold stresses occurring during this growing season which will affect the crop growth and crop transpiration and and production again here so what we will see in this case is that we have a severe reduction of the harvest index to 27% and that is basically imposed by our setting yeah so we have to be a little bit careful with the interpretation of of this value it looks like a high value basically this is imposed by how our manipulation of the crop setting by extending the maturity time with the 500 growing degree days and in the absence of good field data we have to be careful with with these these numbers now we have a final maze farm maze ham and here again a third variety of maze is used we again used the the same approach we need how to impose a fertility stress of 46% and a weed infestation of 15% but we get a much better fit between the observed canopy data also the statistical values are much much more reasonable which is interesting because this was the oldest variety of maze that was used and so in this case this we can say there is a better fit between the default aqua crop settings for for maze and this old variety of maze that is being used in in this farm if we want to do a proper simulation for the other two farms we would actually have to do a more data intense calibration and we see that we get quite a good match between the simulated yield and the observed yield as well for this case now what are the options in here again we have quite favorable rainfalls at times too much rainfall occurring which leads to a reduction in transpiration but there is certainly no severe water stress we have to deal also with reported pronounced period of cold stress which reduces the growth and the yield but in this case compared to what we say as we are now not facing severe stresses during the the second half of the of the season in a case like this it would pay off to fertilize the land because seemingly we have enough water during the remaining half of the growth season and in this case what you will see in terms of the harvest index simulation we see that there are no severe stresses occurring in terms of water that affect the harvest index so the harvest index remains at 48 percent and the major reduction in production and yield is due to fertility stress and here we just give a slide summary and overview and I think you can look better into the details of these of the five simulations from from the powerpoint file later on but I think to stress well if we're interested in the water productivity in terms defined of the kilograms of yield the cubic meters of evapotranspiration we normally would expect a higher water productivity for for mace than for wheat and in this case we basically see that there is no no big difference in fact in some cases the mace is performing lower than the wheat and that is primarily explained by the soil fertility rates where the soil fertility stress in in mace it's much more pronounced than in the cases of of the wheat and as long as you favorable range throughout the entire growing season then you could address those stresses with with fertility management but in the case as in the farm farm a or also in the wheat farms you see where there is still pronounced water stress in the second half of the growing season then adding fertilizers will only have a limited effect as you basically increase the water stress further in the second half of the growing season but working for the and I think the what we say with this application of aqua crop it works quite quite well and quite nicely if you can combine it with field data sets with reported yields and you have some indicators of canopy development during the growth season are quite pleased with how the aqua crop settings for wheat work and with mace clearly we have a more complicated situation it seems to to work reasonably well for one variety but if we are dealing with other varieties clearly the crop file settings will have to be adjusted to reflect the growing dynamics of those varieties better but to do a proper calibration of those crop settings we really would need additional data to do that but I think it's interesting to clearly shows that there is an interest to develop those few the calibrations for different varieties of mace as we're working with the yeah I think this is already discussed as we're working with the water pay project we're also interesting to look at and see we can apply the aqua crop today very different or a very detailed simulation of the production in a field but how does this compare with the Vapor analysis and the water Vapor database so what we've done for five fields or four fields in this case one field wheat farm you see here on the left 40 hectares and three mace fields of smaller scale with one to two hectares and I think the largest of that one is four hectares to obtain also the Vapor data for that same field the same crop and that same growing season and how do that do those data compare then with our simulation results in in aqua crop so from the Vapor Vapor database we use the level two data because that's to do with where these fields were situated in in Kenya so we're dealing with a hundred meter resolution geo reference pixels we applied the the same planting and harvested dates as in aqua crop to define the simulation period or the data period for for Vapor we abstracted the seasonal Vapor transpiration the net production function which is based on the light efficiency which we then converted into above ground biomass for wheat and for maize and then Vapor mostly applies or can be applied with the fixed harvest index to obtain the yield and here in the stable we see the results coming to side by side and first for the wheat farm I think it's interesting to look in here we see immediately okay we have an issue in terms of the simulated Vapor transpiration for aqua crop about 330 millimeters for the growing season whereas from Vapor we only had 200 millimeters 199 so considerable lower there is some form of discrepancy between those two data where we need to look for further detail what is what is happening the same we see basically in terms of biomass the biomass reported from from Vapor is lower than the one simulated in aqua crop but given that the evapotranspiration is much lower than the one in aqua crop you will see that in terms of water productivity there's still quite a difference between Vapor and and aqua crop for for the wheat when we look at the maize situation this is becoming more pronounced and it is we're looking first at the ETA data although there are small farms much smaller farms actually we have a very good comparison between the evapotranspiration from aqua crop and Vapor for all three farms they're all within the five percent margin so that is what we would consider as a very good match between those two types of data where there is a very large discrepancy is between the reported biomass where for instance here in farm b we have a simulated biomass in aqua crop of 14.7 tons per hectare where is Vapor returns a biomass production of 41 tons of biomass for the same field which is actually higher than the maximum potential biomass reported by by by aqua crop this is consistent we see this very large difference in reported biomass between Vapor and and aqua crop which then immediately has an effect on the reported water productivities which then provide very different numbers here 1.12 for the aqua crop simulation and 3.13 for the for the Vapor simulation so I think clearly we this is an issue that we we need to look further into into detail and I think that the main issue that we're dealing with and that we have to look more careful into into detail is how does the Vapor calculation and the simulation of the net production how does that account for a fertility stress especially as in the wheat cases we're dealing with severe fertility stress we know that the canopy development is is is affected it's not only the size of the canopy it's also the the greenness of the canopy and the photosynthetic efficiency of the canopy well thank you maria and thank you herardo those were really nice presentations some of us are new to aqua crop so it was nice to get an overview of of how the model works from the different input parameters and then to see it actually apply to a case study and talk about the you know specific issues that come up when you're applying models in real life and then as well the comparison to the Vapor data and the the work that is being done in that perspective and we will also put up those references and the materials for the the aqua crop on the project website so people who are interested to learn more about how to use the program you can also access the links there so now let's start with the question and answer period we had a lot of nice questions in the chat so we'll put them up on the screen first question is from david hi maria and herardo we all know that for simulation to match with observation good input is required what are the variables that you think must be good to get good results it will be effective that intent aqua crop is able to to to simulate it in a very dynamic way i think we're we're we run into an issue now with with Vapor we have to look closer into how can Vapor in the simulation of the net production function simulate these fertility stresses because if i look here at the previous numbers these very high biomass production numbers derived from the net production function to me indicate that these stresses are not adequately covered yet so that's something that we for the next month are going to look into into more detail to see if we can address that that issue more specific um yeah this is basically we come here to to an end of of the of the presentation and hopefully i can open the floor to some questions and answers later on i think it's important to know that aqua crop is is a free software available from the feo it put a lot of effort in in the valuing net there's also a lot of reference and training material and that you can freely access so here we have provided some further references for you if you're interested to to look in that further uh and these are some of the uh these are the references that sounds like a lot of data is needed to run this program in the presentation for maria how is soil like managed in rain fed farming in our case we we didn't include any soil salinity stresses but uh i don't know maybe here are the ones to to discuss that more in general or to facilitate the discussion um yeah i think i think that's a very good good question i think what our simulation uh shows in here is that we we obtained uh you need good climate data uh daily climate data from a from a weather station in this case we used them from the timmo or tamo uh uh network which is spreading out automatic uh weather stations at secondary schools in uh and primary schools in in in kina so try to locate a uh a good climate uh station as close uh by you need a good soil uh data so you need to know what the soil data is and some basic uh uh canopy cover uh uh data as as we showed in the in three out of the five cases uh if you can find a match between the uh the yield that you record and the canopy uh over time over calendar days uh that that is uh uh from simon with regards to a wheat example with data requirements by top dressing later on in the growth stage but also in two nice cases where we really say okay now we are dealing with a variety of a crop that is really uh acting different uh uh uh than uh as defined in aquacrop then you really need a much more careful uh recording of uh uh the canopy cover during the time of the growth but as well as a recording of when is my crop going from one growth stage into the other growth stage uh so that you can match those uh data uh uh better now i i uh again i think that's an option that uh uh aquacrop provides as an additional option that you need to uh to deal with if you are uh facing sublimity stress we know that it will affect uh uh transpiration of a crop and it will therefore also affect the amount of photosynthesis and the amount of biomass that uh can uh uh can occur normal in situations where we're really very pronounced uh uh rainfall rainy seasons as we're dealing in in Kenya we have a good group image uh next question is concerned with salinity today if you are concerned with uh salinity and these are issues that you need to uh uh take into account them as an address aquacrop provides you with the opportunity if it is in rain fat then you will have to take soil samples uh uh to uh uh to know what is the salinity levels in uh in your soil uh and to set those parameters uh within aquacrop uh yeah it the and probably uh that might have uh have an effect but it will not have the same effect so can you please explain what it meant by water stress soil fertility stress and irrigation stress uh because you already have a reduced canopy size and a reduced uh production level at the earliest stage uh of the of the growing season and I think the whole point I think that is that is clear and that is also with the what we know also from from practices in uh in rain fat conditions uh fertility management and it pays out it's an expensive uh input uh for for for many farmers uh it only pays off uh if you have a secured water supply later in uh in this season and dealing with the rain fat condition that entails a risk yeah and so uh it may not always pay off completely uh from uh uh uh from that point of view to invest in very expensive uh uh fertility uh uh uh to have a very maximum uh yield if you don't uh if you're going to face uh water stresses later on no and that is so that is complicating uh uh the factors uh as well as that's what I mentioned also uh uh by uh in the maze where the actual hardest thing is after uh the maturity of the crop so there is additional time uh where the crop is standing in the field and maybe uh of risk of having full pests and uh and diseases these are not issues that can be uh uh uh uh addressed in uh uh in the aquacrop so that you need field information and then you can say okay I can expect what's the real past of the disease I can expect some different terms for everyone and a question from between the simulation result of high productivity when considering African conditions four to six tons per hectare which is different from data from FAO staff is your observation data from actual farm or from experimental fields and how can we approach that actual yields that farmers get? Think in a in a subsequent comment. Water stress basically uh there's too little water available for the crop uh uh uh uh uh uh to to transpire uh and at that moment uh when there is too little water uh the stomata will close uh and biomass production will immediately uh be uh lower uh and affected uh and uh well in the case of aquacrop we also see that mild water stress so there is still enough water uh for the uh for the crop to use in terms of transpiration but there's not enough water within the plant to be very rigid the crop has water stress it will become uh more uh less rigid uh which may hamper the the rate of canopy growth the the the the the velocity with which the leaves uh will uh will grow so fertility stress is really the uh the lack of nutrients so a fertilizer application uh are there any attempts to replace and that directly affects the state of canopy uh uh and especially also the the greenness of the of the leaves with that how efficient it is in producing biomass aeration stress are those conditions when basically there is too much water and when there is too much water in the field so we have uh uh standing water on the on the field a crop will have a stress in uh uh in oxygen the roots will not be able to uh uh to take up enough uh uh oxygen and we see that this also results in a reduction of transpiration uh and so it's it's very similar exact too much water has a similar effect as too little water and it will again affect the amount of biomass uh that can be produced at that moment now so these yeah and I think it's uh uh so the the files that are really uh uh derived from very uh uh from from the statistics that are provided uh to to FAO by the Kenyan government or other government so and they are uh administrative averages uh uh uh reported so the yields that we have uh used in here and that's also what we say the observed reported yield are those uh yields uh as reported by the farmer so uh and so there will be always be uh uh differences uh and I think in this case uh for commercial settings three to four tons uh or six tons uh uh uh uh uh and depending on uh on the range uh uh that we have also in Mays uh six down is uh is obtainable I think this is uh and another question from David we're looking at it and I think the uh the NDVI that is used in in in VAPOR uh basically you could try to use that also as a as a canopy cover uh data and that's basically I think a similar approach has been used in our simulations for aquacrop although we we took a uh a higher resolution image at 10 by 10 meters uh uh to uh monitor the canopy uh cover 80 80 zero uh values I think this is something that we need to uh do a comparison with uh uh especially to look at for instance in that case of the wheat farm where we saw discrepancy between the reported actual evapotranspiration of uh uh aquacrop and uh and VAPOR and I think it's worth about to look and try to do a bit more comparison okay what are the ETD not files for for VAPOR uh how do they compare to the the weather station files that we are using uh in this case because if there is a large discrepancy in those two climate sets then of course we will also expect deviations to occur in terms of uh of the production functions um we have a question for twitter how do you basically that has uh that can be done with uh NDVI over the ETL uh calibration and the net production function I think that's clearly the issue that we need to to look um in two more detail uh how the net production function is uh is dealing with is uh a stress situation like uh what sort of fertility strike and uh how we can approach that in a different way it's already the the outlines of those plots and the the resolution of the images so there's a mismatch and a challenge with the resolution of VAPOR and the fields of smaller farmers yeah I think it's uh uh and when it's raining it there's cloud power there's a lot of missing data during those particular periods so there's more uncertainty around the values and then there's also a question about the multi crop farms um we already showed in the previous webinars that if we want to use VAPOR we need a lot of field data uh and we can do something with like monoclops but mixed field crops uh is is very challenging to make any estimations and comparisons with the yield data from aquaflip or the fields so those are the the challenges in small water rain threat conditions so as as we can see there's a lot more work to be done by by our projects and by other researchers so it's a really interesting field um that will continue to develop in the future so I think that was our last question and I would like to thank the presenters thank Maria and Gerardo for uh your nice information that you shared with us and I would like to share everybody that attended and was very active in the chat it was nice to see people asking questions and and even answering each other in the chat aside from the answers that we saw in the panel so thank you to everyone and as this is our last webinar in this master class series I would like to thank you all for attending over the last six weeks and for engaging with us on the website as well as the twitter account and we would like to keep you guys engaged so please remember to keep in touch with us um connect with us on twitter follow us on twitter and on youtube and stay in touch via the website we'll be putting out publications as they get finished for different case studies and different information that will help you guys work through water productivity challenges that you're facing we do have more activities that the project will be working on we're actually working on a MOOC a massive online open course which uh in the upcoming months we'll make available to people to use so please keep in touch with about that we'd be interested to have you all attend that as well and we may be doing more webinar series so in the survey that people have completed you gave us some ideas some really nice ideas on how we can continue to engage with you guys over the upcoming months especially since we're not allowed to travel very much so having staying connected via the internet is is really important so again if you have not filled out that survey we really appreciate it it helps us a lot to figure out who you are and the interest that you have and how we can uh morph the project to to help you guys with your concerns and issues so if you have not filled out the survey as soon as this webinar ends it will send you directly to that link if you already filled it out thank you very much we only need you to fill it out once and then we get the information um if you are new please fill it out to help us so again i would like to thank you and i would like to thank abraham uh from mason mason he's really been helping us behind the scenes and setting everything up as well as the other colleagues in the water pit projects and uh our funders