What vegetation indices can tell you about your crop at different growth stages

uh they uh they absorb a lot of energy in the visible spectrum and that's why they reflect the green energy which is actually in the green spectrum and that's why you can see those waves because they are green however if we go here so you can see that the green color is actually in visible uh region of a spectrum however there are like mirroring frayed radiation that cannot be caught with like with our eye however satellites are working pretty much the same as a human eye and however their sensors are able however their sensors are able to recognize this near-infrared radiation and the logic here is pretty simple the healthy leaf that you can see here it reflects near infrared radiation a lot and absorbs a lot of green red and blue energy and the healthy leaf shows us like big difference between near inflated and red radiation when when we look at the sick leaf or that leaf we see that there is like small gap between near and created radiation and red radiation and it's pretty means uh that plant absorbs a lot of this near-infrared radiation and it means that it's sick or that that's something uh that our indices showing not our one but every once yes like uh in dvi in their india and others that's how they work so you can trust satellite data and from here we are moving yeah rim could you please tell a bit more uh about have individual indexes k how indices can help us and uh about in jbi in particular right yep sure so um basically you know there are so many indices out there and probably as you know the most commonly used one is ndvi which shows you the overall healthiness of the crop we're going to be digging in deeper into each index the one that we have in the platform and also a little bit well a few more that would be useful to you also well to analyze your crop but basically in general all vegetation indices um are used in order to understand what's happening uh in your field uh to understand you know how your crop is behaving how it's growing starting from the plantation you know season down to harvesting so we can actually understand whether there are certain problems that struck you know the crops so it could be for example pests disease fungus also the problem we are able to identify and also we can understand you know the dynamics what's going on in the field we can analyze this data not only let's say for this year or what's happening at the moment but also for the past five years so you will understand you know in time um how everything sort of is growing or has been growing in order also to plan your activities in advance that's also very important um in addition by understanding you know the behavior of your crop you're able to estimate your yield and that's of course extremely important and also we are able to uh with the help of the indices along with the analytics that we provide such as you know soil moisture uh weather data like cold stress uh heat stress all of these data you know with the combination of indices they are able to give you sort of the full picture from a to z on uh the state of your crop and what you should do what you should not do in order to preserve the crop and make it even healthier in the area or in the fields um so today's um as i mentioned today's plan is to talk about the four indices that we have so we're going to be talking about ndvi which i mentioned sort of it gives you the overall healthiness of the crop we're going to be talking about ndre and some specific use cases in order to understand when the crop um is ready to be uh well to be harvested we're going to be talking about reiki um you know how to use it in order to understand which problematic areas uh there are you know in the fields and uh yeah and then we're gonna conclude with msavi so we're gonna be digging in as i mentioned in detail uh just in a little bit and i think now perhaps uh well i think we are ready to jump straight to the platform uh but before we do that i think we're gonna turn off um our camera so you guys can have you know can you know see the picture see our screen and also you know for better quality but we will try to answer your questions and of course so we are recording this because i know this is very important for some of you also to share it with your colleagues later on so we're going to be recording this and also posting it on youtube so we'll share the link after the the webinar ends yeah so yeah i think we will switch off the uh videos and we'll begin straight to well jump to the platform so all right all right um and especially that will be useful for australia and canada that are sleeping oh australia is sleeping right now all right so um i believe you can see my screen right yep all right so uh let me then do the following i'm gonna jump straight to the platform so for today's webinar i have decided to show you two different plots they're different one is in canada alberta and the other one is actually in europe where corn is grown so i picked these two for a reason i'm gonna you know tell you a little bit more detail just a little bit later but uh just to go you know to start from the beginning we use data from sentinel 2 that offers the resolution 10 by 10 meters and as i mentioned you will have access to the data from 2016 so you have access to well five years of data in order to understand what's happening you know in the in the field so in this scenario we have you know a field of 126 hectares we have spring barley and the sewing date was actually pretty recent so as you can see uh by the end of march we just planted this crop and the ndvi so let me just remove real quick the data here so it makes more sense for you guys but basically what we are looking at is the ndvi and as you can see it has been pretty flat um throughout this whole year until well a few weeks after we planted uh the crop we start seeing sort of you know uh the ndvi starts increasing so that tells us that the crop now you know is developing in the area and if we go back you know uh in time then for example if i pick some dates from february or from january then i will see something as completely red and this is the scenario so this is from february 7th and as you can see the entire field is completely um red so just sort of to give you an idea you know or a background about nddi so ndvi is the normalized difference vegetation index it shows you as i mentioned this is the overall state or the healthiness of your crop so by understanding the reflected light waves that are emitted from the plants we can understand you know how healthy it is and uh lena also mentioned earlier so the green sort of areas when you're looking at this current picture now the entire field is green so that indicates high level of chlorophyll and that means the plant is healthy it contains a good cell structure because it's you know it's actively absorbing the lights the red lights and it reflects that uh near infrared and vice versa so if the plant is um not doing you know so well then we would see something that's completely red so red indicates two scenarios the first scenario is as you can see this area is completely red uh and like from the sewing date well from the beginning of this year until april the entire area is red so that means it's open soil so ndvi as you can see here by clicking on the legend we can understand that ndvi uh value ranges between minus one to plus one so um areas that has snow you know sand or bearing rock usually would show very low vegetation uh you know value like 0.1 or even less sparse vegetation for example if the area contains a little bit like let's say um you know greenish color a little bit or maybe orange uh then the value would range between let's say 0.2 to 0.5 and anything from 0.6 and above would indicate you know dense vegetation that you know the crop is good it's doing it you know it's doing very well in this scenario for example when we are looking at the current image in february then um well we can look just at the natural color in this scenario because this is in canada and as you guys might know pretty cold in canada uh in february so that would indicate that this area as you can see has snow so obviously this is one of the downsides a little bit when we are using satellite data or when you're working with ndvi so if the area contains a lot of snow the line would be pretty much flat because we are not able to decipher uh the values or you know to understand how the the crop is behaving because it's covered with snow so that's why the area here is pretty much pretty flat but again you know starting from april and until now the field is doing very well as you can see here on the 26th of july now the other field which i showed you earlier which is corn the scenario is a little bit different so that's why i wanted also to show you this field this one we have corn as i mentioned the sewing date was in the beginning of april and um as you can see here the scenario does not resemble like the one in canada and that is actually there is a reason behind this so we just planted this in april but the ndvi kind of was not flat as it was you know in the field in canada because if you for example click on the 14th of march you will see that the area is orange and if you remember when we have something uh like you know sparks vegetation when the value ranges between 0.2 to 0.5 that that means that the area you know contained wheat and that's why the nba was not really flat but it's kind of like we we have some activities going on but that is actually wheat in the area and if example you click on february then of course we will see something completely red because that would indicate open soil so that's why it's very important to put the right sowing date and also to know the kind of crop that you're dealing with in order to understand the tendency the you know the tendencies and also to interpret the data that you're seeing in the platform um yeah so i would say that you know ndvi in general of course um you know is a great tool and a great indicator to show you know the health of the crop uh but and again of course if there are some certain you know areas where let's say some there are some red spots hypothetically speaking uh let's say there are some red spots here and there in the field that that would indicate a certain problem and of course the earlier you identify those red spots that could indicate a disease or pest or fungus the more you can maximize you know the canopy size so that would give you know higher yield at the end and this is what you know everybody is aiming for um so yeah i would say this is uh for ndvi um now um uh just one question from me here so you were saying that farmers should check in dvi level at least like every three or five days you need to follow the changes and i'm wondering if i'm a farmer and i have like hundreds of fails or even a hundred fields so do i have to check at all of them um sure that's actually a good question so the thing is uh we have a lot of clients uh and each you know client you know has fields scattered across let's say brazil or argentina or you know other parts of the world of course you have two options here either you can do the old you know fashionable way where you can just go to every field every single day and check the state of your uh of your field or you can go to our field leaderboard tab and in the field builder board basically we will be able to see uh the fields uh all the fields um and you can filter them out by value change or by index value by crop type by group type so you can of course when you're adding fields in the platform you have the option to uh you know categorize each field with them maybe you know in different countries or uh you know in in different categories it's up to you to decide really but basically we will be seeing something that uh for example if we have here sudden drop in value so brett this is something that you will need you know to uh to check definitely so you can click on open and the platform will take you straight to that field and you will study uh the field in detail in order to understand what's the the reason behind the sudden drop let's say for example of 0.4 so that could indicate perhaps really um maybe a certain problem that really struck the field and would require immediate attention but of course if you're checking some green values like these ones this is something that you would want to see definitely so that would tell you that yes you know the field is doing really great and uh there is also actually another thing so you get email notifications about these uh you know the sudden changes in values regardless if it's a plus or minus so uh the notification email uh will be sent and we it will show you that well first of all it's gonna show you two pictures the most recent picture and the one before it prior let's say to the one that's been taken let's say two days ago so you're gonna see the changes if it's a plus that's a good thing if it's minus again you will need to go and um check the field in detail so uh yeah that's for ndvi i would say but now uh probably i will talk also about ndre just real quick because uh lena will be covering a little bit more in dre in detail but basically ndre is uh will allow you sort of to detect the very different variations in the health let's say of crops again these indices could be used for different crops for example cereals like wheat spring barley corn maybe oil seeds like sunflower cotton plantation crops even tea ra coffee rubber or even other uh crops uh like we have a lot of clients you know who work with sugar cane so it could be us also for sugarcane and tobacco but uh yeah so for ndre as i mentioned um it's able to detect these different variations especially in the later stage uh it sort of highlights you know the shows you the high biomass crop and also permanent trees crop as well if you're using them um so we can detect you know the changes in chlorophyll level content within the leaf itself um so that's why of course it's very important to look at ndre at the later stage of crop development um so yeah that's for ndre but perhaps you know a couple of comments also from my side before i you know give the floor to lena um i would say that if you know guys if you're working with thick permanent crops and perhaps some other dense crops or even the you know later growth stages for some specific ones then of course i highly recommend to go with ndre but um let's say if you know if your crop transition from let's say the seed or growth stage to thick canopies in one season then i would highly recommend using ndvi and ndre at the same time uh but yeah ndre basically also is used for i would say more on the intensive side for managing applications throughout the growing season for them crops uh but yeah i think this is it i pretty much covered ndvi and ndre um lina would you like to take over and tell us more about and savvy and frankie right so let me share my screen and i've been following all your questions room we have a lot of questions oh boy yeah so i'll try to cover some of them during during my uh speech right now talking about mscbi and other indices and afterwards we'll we'll have a minute you know to answer all your questions because they look really interesting yeah so msavi let's start from the very very vegetation uh beginning from the very beginning like from the vegetation beginning and uh as you know msavi it is called modified soil adjacent index and it is specially designed to minimize the soil implants on canopy spectra easy like saying in other words any cvi only considers canopy changes you know it it doesn't consider soil and it shows us deviation within the canopy only so msavi values are usually higher than in dvi and we use that at the early stages of vegetation development it's like one or two months after sewing and this particular case we have here this is winter wraps that which which was sewn in august and to let's go back to previous year the platform allows you to get historical images here from 2019 and five years back so we'll select just the previous year to see what was happening there so you can you can go back and while we are they waiting [Music] right so some of you are asking about the the image frequency so uh we got it depends on on the region where you are located because the satellites like goes all over the whole earth and it might take from 10 it might take from three to seven days depending on your region and sometimes where uh images are covered with clouds more than 50 50 we don't get them into the system that's why you might have what um i believe because the cloud masking algorithm is working in the background right the one that we have uh in the system so that's why we don't show the imagery yes yes right so uh that's why like having our own clown masking processing we don't get those images uh into the system because they're like useless for analytics okay so getting back to our msavi index uh the crop winter wraps it was sewn in august and let zoom a bit in to see what was happening there in august uh we'll switch off the soil moisture for a moment in order not to confuse you and here let's select the closest image within two months after sewing and we're selecting october 15th here so it's been it's been a month and a half since when and what we see here you can see that uh in dvi showing like the red values very low values and it's something that confuses you it's not clear so okay what it means for siblings and then here we are switching to msavi to minimize the soil influence and here what we see is that in the northern part of the field the color is actually green which means that the development in the northern part really differs from the development in the south part which is like more uh which is more whitish here and right so and you can see that the whitish part takes about 70 percent of the failed and it's a lot and we can understand that the the field is actually heterogeneous and it's something is happening the siblings are developing in the northern part better so what might be the reason and if you pay attention you know to the analytics below and we can notice that from august until september there were no precipitations at all and also we can uh click on root zone soil moisture to see that the soil moisture level there was critically low it's like 13 or even 12 percent and as you know rhapsod is really sensitive to the soil temperature and soil moisture especially during sowing period so this is something that where you should pay attention so that might be the reason of poor development of the seedlings right so uh the client in this case he detected these changes these deviations at early stage and they had a chance to resolve their crop so the platform allows you uh you know to detect changes at the early stages and do something with them so that's it about msavi and now we're switching to india switching to inter e right yeah um lena before you um you know jump in um about the other in this i just wondering so um is there like a normal value when we are looking at different indices for crops like you know winter wheat or any kind of crop out there is there like a certain number i can i should stick to uh to know that my crop is actually doing really good right uh very good question and i believe some of you also have this kind of question there in the chat you were asking so if i see you know ndi value which is all point a what it actually means for me um i can answer this you should understand and i want to highlight the attention here that indices like in dvi in their e they are created uh with com for comparative analysis so it doesn't actually mean if you see like 0.6 0.4 value there it actually means nothing if you don't if you don't know your field and you can only tell from the image like here like i was showing you that there is some difference in vegetation but what it means we can tell it so uh this is something that you can gain uh using deductive methods like going into the field opening your ndvi uh images and compare so okay i have here 0.4 what it actually means for my crop is it like winter wraps it corn wheat what it actually means for the weather conditions that i have there for the type so for the soil type that is there uh so if you're just like really new uh to these kind of platforms you will need to spend at least one season you know to understand what those values actually mean for you and for your particular crop so this question is something that only you can answer them considering your weather conditions soil type and uh activities there perhaps irrigation if you're if you're doing this yes so this is very very good question and switching back to india let me let me explain here so in dvi for your better understanding ndvi uses red uh spectrum and near-infrared and red spectrum it penetrates only the canopy and it can use and it can show the picture only in terms of the canopy when in dre it is using um red age uh radiate channel and it penetrates deeper than the canopy and it can show you can give you more understanding about about the real biomass presented on your field and in dvi and in their e they are using simultaneously so uh here in this field we have winter wheat uh and we have image from june 27th so those indices are used uh during um in the middle of the vegetation development and if you are switching to ndvi i want to show you the difference so if you if you know your field and you see that in dvi showing not really like the real good picture uh or it's not real picture as you see directly in your field you probably need to compare it with other in the test so here we can see that ndvi shows pretty good picture right and when we switch to india e like except this uh spot here in in the center and we switch to endear e and you can see that uh in comparison in comparison to higher vegetation we have like poor one in the middle it is good but it is worse following these vegetation changes in the next images so we go back for example not go back just we go to july 5th you know to understand whoa okay did my field changed and what we can see here in the air e actually showing like homogeneous values and it doesn't look like truth it sometimes happens because indices are really sensitive to um some noise you know atmospheric noise or something and you should consider it so in this case we can switch to red age chlorophyll red age chlorophyll it shows you the chlorophyll um amount presented in in the plants and uh it sometimes can be different from from ndvi and in their e because it uses uh it uses channels which are like designed to understand the content of the chlorophyll in the field and what we can see here the picture is totally different red h chlorophyll showing really low values and chlorophyll is directly connected with the nitrogen and that's mean that probably means that your plants are lacking nitrogen yeah so real quick here note um so as far as i know like around 30 uh or approximately 30 of the total nitrogen is always or usually applied at the beginning you know of the crop let's say when you have the sowing date or as soon as the crop emerges you know you apply that 30 percent content of nitrogen so does that mean in this scenario that the nitrogen level was not like if we or the client or let's say the farmer of the steel did not apply the right amount of nitrogen and that's why they started having problems in this field yeah that's this is a good question yeah it might be the problem because we know that nitrogen is applicable at the early stages because it uh they test better with plans that time uh it might be the reason uh and also also you know that nitrogen level falling down after heavy rains so i would recommend you to use red age chlorophyll uh after rains like checking the level and compare it to in dvi and india in in their eel values and also uh we yeah i remember this question about variable rates we do have a variable rate application map where you can build them and for nitrogen you can use red-h chlorophyll you know to build your maps for nitrogen variable application based on radiation chlorophyll yeah so that's it about about indices um i think lena we have some very interesting questions that i think maybe you know we can just take a little break here and answer some of them um so let's see the first one or the most i would say commonly used question is about soil moisture and uh yeah you guys have asked us how do we calculate so if you see here in the platform we have two levels of soil moisture we have root zone level and we have surface level uh usually to calculate the soil water content we use a system of interconnected moisture and heat flow equations so basically we look at data like historical data recorded and also forecasted so we have access to historical data as lena said for the past 10 years plus we have access to forecast data for let's say for the next 14 days so we look at this meteorological data air temperature humidity amount of precipitation and even soil parameter and then we are able to model that using our you know equations and display the soil moisture content now um the surface level is usually down to as i mentioned to five centimeters and root zone is down to 60 but now root zone is a little bit different um so it's calculated by taking or i would say the amount of water in the roots or considering i would say the plant transpiration rate we take also into consideration root length and also the current soil moisture that we have from the surface level and that's how we are able to uh model soil moisture for two different sort of kinds so we have surface and again root zone so i hope that answers um soil moisture question um yeah maybe we know i'll answer the next one uh i wanted to add here so in case you are in that region where you don't have like weather station close to your fields and it's not showing you know the real picture the soil moisture we're taking from radar satellites so it actually measures it there and you can rely on that uh in case you don't have like you know good source of weather near you yeah and uh answering the question that i have here um yeah can you see use this to see with pressure yeah that's something that i didn't cover uh with the msavi actually msavi you can use uh after harvesting to track uh to track weights there or like you know in early spring for uh spring crops like to detect weights there right yeah uh probably i'll take the next question um so this question is from einar i believe um so do government industries and suppliers have the same access as the farmer um yes well actually you know when you're using our platform everybody has you know basically global access to all the fields around the world so you can basically monitor all sorts of fields of course this will depend on the project so for what we are showing you right now is something that is like out of the box um if you are uh you know you want to do for example crop classification yield prediction of course that kind of data would be uh well classified you know that will only be available to the customer that you know requires these kind of projects so if you have these kind of projects that you know you want to know maybe estimate yield prediction or yield forecast for example for uh corn or sugar sugarcane or whatever it is then of course send us the details the details at sales.com and of course we'll try to come back to you with the details on how you know we can work together in order to provide you with the with the yield for example estimations um yeah and maybe another question or um yeah concerning um sorry swishing one on um concerning this question about governance industries and suppliers uh yeah like um the thing to add here uh if you're talking about the data from the data uh data processing you hear like no one has access to to your data that you are processing there to the records the records that you leave there in the platform everything is separated so if it is question about like you know data protection or data processing yeah like you have different totally different accounts running on different amazon servers which are protected and like you are the only person who has access to your data and yeah i wanted you know quickly uh you were asking about variable fertilizer rates and that's something i wanted like to show you uh real quick uh here here like we have a tab which is called zoning and that's where you can you can build your variable rates uh maps based on a single image which is usually you know enough for nitrogen application or something so you just select uh the index that you want to analyze and select like the image uh cloudless one and select the number of zones and you just press calculate and we have like you know different zones there so this one is working for nitrogen uh variable rate application and also we did have productivity maps the productivity maps are built based on you know several images through the season and they show you the quality of your fail it answers your question whether it is homogeneous or more heterogeneous do like do i have to apply variable rate applications there and everything so first time it usually takes time to calculate it but yeah that's it answering the question about variable rates and i think we'll answer other questions later yeah and now okay ream could you please help us to sum everything up um sure so basically uh we you know we have uh showed you four indices basically uh the one that are available in the platform uh we will be covering uh a few others in a little bit uh but basically be so in order let's say to know which cup to begin with or which index to use in order to understand you know or where to start let's say right so you added your field and you don't know which index to start with so my recommendation is to start with msavi so msa vi so which is this is very useful in order to understand the earlier stages of plan development and then moving on you can start or you can check ndvi again just to show you the active sort of stage or the growth development how you know how the field or how the crop is behaving in the field then you can move to ndre and of course it's recommended it's highly recommended to use ndre along with ndvi in order to understand the sort of layer growth stage development of your crops um and yeah and of course reiki uh to monitor just say the chlorophyll content uh just to show you the the signs you know of how healthy the crop is in the field uh we'll be covering indices like evi and others just in a little bit but before we do so i just wanted to bring your attention to something that we didn't discuss earlier which is as you can see on the platform we have something called growth stages um so maybe lina do you want to do you have yeah i can share my screen to show it yeah so um we have also growth stages but we are not going to be covering uh all the details regarding rose stages but as you can see here we can understand uh at what stage the crop is at the moment so uh for now we have growth stages available for i would say up to seven crops including soya beans corn spring barley um wheat uh and a few others in selective countries uh but basically we are able to understand uh cotton two yes it's included uh but so we can start by understanding you know how the crop is behaving uh starting let's say from leaf development delivering booting until fruit development um but of course maybe some of you is wondering like how this data is calculated so growth stages we just look at the overall you know we so we have our own you know engineers like scientists and phd doctors who study you know uh the each crop type how it behaves you know in certain temperature under certain conditions and we correlate that data by looking at the temperatures the weather data the active the daily active example of something the sum of active temperature i'm sorry weather data soil moisture also we look at ndvi or other indices in order to understand so in this scenario as you can see in the platform when you look when you click on the first or just over your mouse on the first let's say leave development so lead development started happening just a few weeks after we just planted the crop so here we understand that actually the data correlates with uh what we have of course we're gonna be with the ndvi i mean um of course i think we can't really cover growth stages in one webinar because it's going to take a long time so i think perhaps in the future we can dig in a little bit deeper into growth stages perhaps even soil moisture because i think this is very common these are like you know very common requests so probably we're going to be covering them uh later on but i just wanted to you know to say that we actually uh have this in the platform and we can you know show different growth stages development for different source of crops but yeah so before i give the floor to lena i think you know we talked about the you know the the pros of using um our indices uh but we didn't discuss perhaps the cons i would say and there are a few cons so the first one of course when you're working with ndvi as you noticed earlier when we were looking and analyzing the field in in canada we had a lot of snow but uh the ndvi was not able sort of we were not able to identify what's this you know what's the state of the crop so the ndvi was pretty much flat in the area so of course ndvi as i've seen is sensitive to these kind of noise so clouds snow uh you know sun wave angle and all of this you know um affects uh us you know affects the you know the indices and we're not really able to understand what's going on in the crop so yeah that's one weakness also another one is um it's actually not very relevant when we are talking about yields because you know when the yield depends on small fruits especially for plants that has a lot of leaves so like strawberries so of course in dvi also uh which i would say a con in this kind of scenario um yeah yeah so let me let me take it from here because like related to our next huge topic which is like i i i sorry question about why don't we why don't we use a the e v e i or other indices um we do updates every two weeks so please send us what kind of businesses you'd like to see in our platform and we'll do it for you and now yeah i'm going to cover like topic about alai this is like complex and difficult topic so be ready and i'll try to do my best to explain it and i think we'll have another uh another session about yield prediction where i will cover this topic like uh in detail but now uh just given you understanding you know what is the difference between in dvi and other indices we've we've been talking about and lei so uh the one of the weaknesses of ndvi that is actually not a quantitative measure that's something that i told you uh before it's more about a comparison of vegetation density between fields you know within a field for example where you can say okay this here i have more dance vegetation and here like poor poor one uh it's good to compare it between regions countries so scientists you are usually using it uh to compare the yield performance between different countries like in europe and and the us but this is not a quantity of measure and that's where a leaf area index coming come in place yeah so what is leaf area and why is it important so uh leaves are one of the main plant organs and are responsible for the productivity of the plant and on a larger scale it is like responsible of an of an ecosystem or a farm and and like we should investigate the leaf area index in order to uh in order to predict yield or in order to optimize yields uh considering the changing the constant changing of climate so that's why we we need it yeah moving forward to this slide so what is leaf area index this is the ratio of the leaf area per unit ground area so if you look at this image you can see you can differentiate about 10 leaves that are occupying about 30 30 percent of ground area and lei in this case will be 0.4 and on the next image you can see like like about 20 leaves that are occupying about 80 percent of the ground area it's like lei in this case would be 0.8 so it's like when you're looking at the plant from the above and you can see the number of leaves covering the ground and uh scientists are using li you know to consider uh crop uh to measure crop development at different stages and then they are using those in crop simulation models to know uh the the dynamics of crop development so uh lei is something that telling you about the actual biomass uh in in the field so uh here you can see you know to give you a better understanding so as i told you before ntvi is using red uh spec like red waves and they penetrate only the canopy so ndvi only shows you you know it gives you some knowledge about the plant density on the canopy level in there he uses red age uh channel which is at the end of the spectrum here and it penetrates a bit lower but when we are talking about lei it gives you the full picture you know the biomass uh in the field and it's like quantitative measurement and that's something that you can use in the models uh climate models crop simulation models uh to use to see the development of the plants and uh i'll just i'll be like quick here and i'll just touch those models so crop simulation models you probably heard about uh vofost or promet or this assad and uh those are like simulation models for the quantitative analysis of the growth and production production of annual field crops this is about foster and this is how this is actually the software which helps you to uh simulate the crop development and uh can give you like information about yield in the end so they work pretty much the same so they have pretty much the same logic uh we supply the model with crop uh crop data this is genotype um of the crops sometimes those models require uh some management data appliance fertilizers or uh applying pesticides so some information about management soil data the type of the soil is very important and daily weather and then when the crop starts developing at the different vegetation stages we are uh supporting this uh this model with satellite data satellite data and data in this case means lei so and uh we put this lei index at the early vegetation development and later and later on to get more accurate data you know and here so for example if we apply ali data two months before harvest we'll get about 70 of accuracy for yield prediction if you like apply it two weeks before the harvest will will have about 90 of your prediction sure and actually i just want to mention one thing uh because just considering you know when we are doing yield uh forecasting you know yield prediction model you know for custom projects um it's very also important to acquire some data from your side so for example you know first of all we would need data on the current season so first of all we'll need you know deporting of the fields where you know you're interested in uh doing you know yield prediction uh well you know the you know the accordance the aoi the year the crop type the sowing date which is extremely important the more data we have of course the more accurate the you know the the more the higher the accuracy will be so and again the same uh the closer we are to the harvesting uh season the the more accurate the data is so we'll need also to acquire you know sowing dates uh from your side the soil type if you have of course and the same kind of data not only for the current season but also for previous seasons as well um preferably let's say for the past three to four years so if you have you know these kind of uh requests please drop us an email and we'll be able to do you know yield estimations or yield forecasts uh for basically whatever crop i would say um yeah in any country we'll of course we'll have to work together uh on this in order to yeah make it work right uh yeah thank you rim and last like breathe out a bit and uh continue so this is really really complex uh topic and uh like eos since we have our own research and development team we can help with lai estimation because somebody i i i just saw in the chat somebody asked asked how we can calculate uh how to calculate a lie so um this is not just index that calculates using bands you know like in the eye in dvi rad and and uh near infrared radiation so this is a modeling index and it includes not only bands and it's like complex ones so in case um you have in mind like having your yield prediction project because it can be done on the country region level on the field level and that's that should be a team work like supporting and supplying data from from both sides yeah and just really quick so we had a project with kazakhstan ministry of agriculture and we were helping them with calculating lei so they can apply there in their yield prediction models and since we were talking today not about yield prediction but more about indices let me show you like the real difference between li and dvi so uh we have you can see here images from may 1st june and july so yes lei should be calculated uh through like time at time rate range so it's not something that should do you once um and you can see the obvious difference between li and in dvi so especially at the early stages so lei is more reliable to use in those models so you can see if ndvi um in may is showing like rad values here then okay like in dvi is very sensitive to soil reflectance to atmospheric uh noise uh it sometimes cannot see like the difference at very early stages when lei is like actually designed for that and uh here you can see the difference between those indices and closer of course to the july to the higher vegetation stages they are a bit similar but still we remember that ndvi is for comparison and lei is about quantity yes the the actual biomass in the field so with lei we we can get more reliable and more accurate yield results at the early season yeah so that's it about lei i think we'll cover we'll cover uh crop modeling uh and simulation models later we'll share our experience within eos like next webinar and like we have now to cover oh rim we have to cover now evi and sabi just one minute i know that everyone like i think tired too too much information so what is the difference between a evi and savi uh in comparison to in dvi actually they're using pretty much the same uh bands and pretty much the same formula the only difference is that there is like there they have in their formula a variable which can range so we can specify this variable depending on uh on the re like region on the territory and the density of the uh crop so in in case with evi uh we usually use it to analyze areas of birth of earth with large amount of chlorophyll you know like rainforests uh it works great for a wine yard so you will see i have attached for you um uh research about relationship between evi and grapevine phenology so you can check it so it works great for mine for wine yards and savi is vice versa it is usually used for arid regions for really dry land regions and they have this like variable in their formula and if you have like very dry land uh the index will grow like this variable will grow but if you have like usual uh land with usual crop density this l variable will never change it and you'll have the same values and in dvi so yes if um if you know about those indices if you and you'd like to have them you can email us and you'll get them in the crop monitoring like in several next iterations or of course you can calculate them yourself and working with those variables or uh we have another you know easier way yeah spring right yeah so i can perhaps elaborate a little bit more on this so um i know well actually no no but we work with a lot of clients who are building their own platform or who are using maybe you know your own maybe you're using your own farm management solution or developing a system from scratch where you're combining not only satellite data but maybe um other sorts of uh maybe systems let's say uh track management so whatever it is so um if you are uh you know you need our data then of course you can use our api and you can extract all the data that is available in the platform we didn't really dig in the platform today we just wanted to show you more about the indices um and you know on how to interpret the data um and uh you know the different visuals that we get uh when you're looking at the fields but basically with the api um everything is available like in the platform but there are two additional uh i would say features the first one is you have access to more satellite imagery so you have access to uh not only sentinel due data that offers a resolution 10x10 but also landsat 7 8 modis and others plus you have more indices to work with and as you can see here in the slide you don't uh so you don't only have access to the four indices that we showed you today but also to evi or ndsi for example that is snow index msi and others so you can also pull that data extracted and incorporate it or integrated within your own platform so that's one option or one scenario you can go around or maybe you know you can go with this approach um yeah i think this is it um perhaps you know uh we'll do we have some time maybe to answer a couple of questions yeah i think we can switch our cameras on yeah and answer several of the chests because they're really interesting yeah all right okay um [Music] let me go through uh um okay when choosing uh alejandro like i hope i spell it right when choosing fields for agricultural crops which indexes in the sauce would be recommended uh right so um considering in dvi like as we mentioned individual really actually has some like cons uh and uh it usually works great you know for some serials for really dance crop with good density and with the crop um which like yield performance depends on the leaves so it actually that's why it makes sense you know to monitor this kind of crop and uh allocate some stressed areas there but if you're talking about vegetables uh some kind of vegetables like it it also works for a potato uh but like for for berries uh for trees and dvi doesn't always work for those so yes you can see some changes with in the vegetation and you can allocate some stressed areas but it might probably mean nothing to you because you don't see because those plants are more like street 3d model like trees and it those changes doesn't tell you anything so you any way you need to go into your uh field and check it so uh i would recommend here still you can try using ndvi because it it really helps like you know to indicate some stressed area but you need to try doing scouting go there open the app check what it says like you know in dvi value 0.2 0.4 and check okay what it actually means to me with my crop yeah perhaps um like the other questions because a lot of you guys i've noticed in the chat that a lot of you are asking whether you know uh you can leave you know whether you can use certain indices uh like ar vi in the platform itself so at the moment as i mentioned we offer only four indices however if you would like to see um some index in the platform then we can do that um if you don't see it in the api list but you have perhaps the algorithm then of course you can share that with us or just let us know what kind of index you would like to see in the platform you can reach out to us sales eos.com and we'll estimate you know um how long would it take for us to do this usually uh you know based on uh on what we know um usually takes around i would say two i would say we can add two to three indices a week again depending on where you know uh which country it is or what kind of index you guys want us to incorporate in the platform so yeah please leave us an email right and can you help me with that um um where we can find information on active temperatures related to the different types of crops okay i'll i'll show you let me share my screen all right yeah so you okay uh we have let me select this field for example somewhere in canada we switch to weather data and that's where we're showing more detailed weather analytics and as rim previously mentioned would do show accumulated precipitation and some objective temperatures you can actually specify here the base temperature depending on the crop that you have there and uh it's like really easy to follow the correlation between like precipitations and sum of active temperatures uh considering like the grub the crop stages so and you can understand whether your crop has reached uh the number of of temperatures that is required you know to transport to the next stage growth stage so that's where you can you can compare it you can compare you can follow daily precipitation and daily temperatures uh in comparison to your uh growth stages right i hope i answered this question if you if you meant this yeah uh anything else rain i think we are running out of time unless we take one more question one more question yeah so um i i see that we have a lot of question about about soil moisture and about lei okay whisperer will got two another webinars covering soil moisture we'll show you like uh we'll show you what okay and sharing my screen how how we got this data like how we do modeling based on that on the level like um that is comprehensible for everyone and we'll do another webinar covering yield prediction models you know how we do that in the easy way so you can understand like what if you've ever thought about that so what do we need from your end and like from our end so you wanted to answer one more question i think that's it because there are so many questions and we really know now which ones do you to answer so um we're gonna send the reporting um and the materials that we share today to everybody who attended the webinar um of course um if you have any questions do drop us an email um and i hope this has been helpful uh we'll probably pick the next topic but we're not gonna reveal the topic yet uh but uh well you know we'll do as linda mentioned we will do um you know upcoming webinars for yield prediction soil moisture growth stages for different uh for different crop types in different countries how we calculate that data yeah and of course if you have any ideas then do you know do share with us we would love to hear your opinion on this um yeah this is it uh thank you so much for your time i hope this has been helpful uh we'll see you next time thank you thank you everyone and like if you want you can email your questions at sales at eos eos.com and we'll answer them so sorry if we didn't cover some kind of questions that were really important i think that's it yeah thank you very much thank you for joining us have a great day there night evening bye okay

Channel: EOS Crop Monitoring

Views: 2,972

Keywords:

Id: 8RiNBH4KTOI

Length: 60min 21sec (3621 seconds)

Published: Fri Jul 31 2020