Home » General
Category Archives: General
Raedan Sharry, Ph.D. candidate under advisement of B. Arnall
Brian Arnall, Precision Nutrient Management Specialist
The data about to be reported is from the study we have fondly named “Burn Baby Burn”, you will see why soon enough.
Grain Sorghum production continues to be an important component of many growers crop rotations in the Great Plains. However, for many growers who focus primarily on small grains production, equipment restraints may impose limits on in season nitrogen (N) management. When producers are able to delay the application until in-season it helps to ensure that N is available to the crop at the time of increased uptake during the reproductive stages of the crops life. Producers often have access to equipment and technologies that may be used to take advantage of improved N application timing, but may worry about the negative effects that nitrogen can have if the fertilizer is inadvertently applied to plant material. An experiment was initiated in Central Oklahoma to evaluate the yield response of grain sorghum to in-season nitrogen application methods.
Trials were placed at Lake Carl Blackwell near Stillwater, Perkins and Chickasha Oklahoma and included 9 in-season fertilization methods and a 0 nitrogen control. Treatments are listed in Table 1 below.
In total 120 lbs of N was applied to all treatments receiving in-season applications. 60 lbs was applied at planting to all treatments including the “Zero N Control”. The remaining 60 lbs. of N was applied according to application method in-season. The urea was applied by hand and the liquid treatments a push cart with adjustable boom height (Figure 1) was used to apply the UAN. Applications were made mid day at V8 growth stage. The temperature at the time of all applications was about 90 F and humidity below 75%. Nozzle position for 30″ and 60″ was set for between rows.
At two of the three locations (Stillwater and Perkins) the addition of 60 lbs. of N in-season increased yield above the control treatment. At the Stillwater (Lake Carl Blackwell) location there were no statistical differences (α=0.05) between in-season fertilized treatments except the T-Bar 20” treatment (Figure 2). The Perkins location (Figure 3) provided a similar result in which again there was no statistical difference between fertilized treatments, excluding the T-Bar 20” treatment.
The Chickasha location differed in that additional in-season nitrogen did not improve yield (Figure 4). While we want a response to applied N, in the case it allows use to solely evaluate the impact of burn associated with N application. The T-bar 20” treatment statistically negatively impacted grain yield and the FlatFan-20″ did at α=0.10, which means we are only 90% confident the yield lose was due to treatment. This response has been consistent across all three locations, on average decreasing yield approximately 21 bu/ac relative to the individual site grain yield average.
Even though it was mentioned for Chickasha, it is also important to note that while it was not statistically significant (α=0.05) the FF- 20” treatment (Flat Fan nozzles above canopy on 20” spacing) trended towards decreasing yields at all 3 locations and is likely detrimental to crop performance. At all locations substantial damage to leaf material was observed, similar to that pictured in Figure 5 below. Several of the treatments damaged leaf material on the plant through burn injury, but most were not negatively impactful on grain yield in the 2021 growing season. Grain sorghum yield did not benefit from moving the application point below the canopy using drop attachments, nor did adjusting nozzle spacing from 30 to 60”. Source was not a significant factor impacting grain yield regardless of it application method.
The observations from this study show that many of the in-season nitrogen application methods that are available to growers will not negatively impact yield. This however does not apply to tools such as the T-Bar. Similar tools that concentrate large amounts of N to leaf material are also likely to produce similar results. It is important to note that the T-bar was used on 20” spacings and not tested otherwise. Moving the spacing of the T-bar may lead to different results.
Growers who are looking to move N applications in their grain sorghum crop to in-season to capture the benefits associated will likely be able to with equipment that is already available to them. While leaf damage may be observed under sub-optimal application methods, damage is unlikely to contribute to significant yield loss. However, growers should keep in mind that environmental conditions may have a significant impact on the results seen from these types of application as growers should always look to limit stress to the plant when possible.
We of course will be putting out a second year of this study and will share the results when we can.
For more information or questions contact
Brian Arnall firstname.lastname@example.org 405.744.1722
In the past couple years significant efforts have been made to produce N fixing microorganisms that can be utilized in an agriculture system. The atmosphere is 78% N2 and prokaryotic microorganisms such as the bacteria species Azotobacter, Bacillus, Clostridium, and Klebsiella take that N2 gas and turn it into plant available NH4. These organisms have been around providing nitrogen for plants, for as long as there has been plants. In agriculture we have heavily utilized their relationship with legumes however have struggled bringing them into other realms of production. Naturally they tend to be found in areas that are very low levels of nitrogen. For example, prokaryotes were found in the un-fertilized check of the 130-year-old Magruder Plots but are not found any other treatment that receives fertilizer organic or commercial.
Now there are several products marketed as containing N fixing microorganisms suited for use in today’s corn, sorghum, and wheat production. While I have an active research program evaluating the use of such materials in Oklahoma, this blog will not address what works or how well. This blog will touch upon my thoughts on how to utilize a technology such as this if you pull the trigger to implement.
So there is one key to getting a ROI on products that create plant available nitrogen, and it’s a really simple key.
Under Apply Nitrogen
If you apply enough or more N than the crops needs, then there is ZERO value in a product that creates more N. For example, applying one of these products in your 250-bushel yield goal corn after you’ve already laid down 300 lbs of N preplant. Unless you lose it all to leaching, your probability of seeing a ROI on your biological investment is pretty poor. I have a hard time understanding the thought process behind paying for a N fixing product and not lowering your fertilizer rate. I can see one of two reasons. 1) You believe you historically under apply N and are losing yield because of such 2) Are in an environment which has a high potential of late season N losses, and you are unable to make recovery applications.
So what to do if using a N Fixer? I do not have the confidence yet to say, “Apply X product, it will produce Y lbs of N, so cut your rate by Y lbs”. That uncertainty is one of the greatest challenges, not knowing will I get 10 lbs or 40 lbs? If I did, then I would just subtract that off my planned rate. Side note, as someone who has been doing on farm N rate studies for a decade plus, I would have to add that most were likely over applying by that much and could cut back anyways. For me the use of the N Fixers should force your hand into utilizing in-season N applications, regardless the crop. So that you can better predict or determine impact of the product.
This is where the use of a refence strip (N-Rich or Zero N) is the golden ticket. We need a way to quickly evaluate the amount of N the crop has access to. The N-Rich method works best when preplant N is drawn way back. I would add that reduced pre-plant is a great scenario for N Fixers. The N-Rich in comparison to the rest of the field will provide you guidance towards your in-season goals. If the N-Fixers are doing a great job the N-Rich will not be showing up any time soon and you can make your N rate adjustments accordingly. If you are a Pre-plant or die kind of farmer, then I say you need to pull back the reins on the preplant rate but give the N Fixers some room to add value and add in your Zero N strips. These will again let you observe what is happening in the soil apart from your fertilizer. If it is getting on the late side of in-season N and you cannot find your zero, might be a good time to walk away and hang up the fertilizer applicator keys. I have lots of blogs and pubs on the use of reference strip so send me a note if you want to dive further into these approaches.
Feel free to reach out with questions or comments. B.email@example.com
Brian Arnall, Precision Nutrient Management Extension Specialist.
The basics for nitrogen (N) fertilizer rate determination can be described in a mechanistic approach by the Stanford Equation NFert = ( NCrop – NSoil ) / Neff. This equations states that the N fertilizer rate is equal to the amount of nitrogen taken up by the crop minus the amount of nitrogen supply by the soil, divided by the efficiency of the nitrogen fertilizer used. I outline the importance of this equation in the blog “Components of a variable rate nitrogen recommendations“.
There are nitrogen “Easy Buttons” which utilizes averages collected over diverse environments to create accurate N rate recommendations. The best example of this is the yield goal rules of thumb such as wheats 2.0 lbs N per yield goal bushel minus soil test nitrate. Yield goals are generally calculated as the average of the best 3 out of 5 years, or the 5-year average times 20%. Also, the 2.0 lbs of N is more than what is in a bushel as it also adds in an efficiency factor or a 0.5 lbs per bushel cushion. This method and others like it provide an accurate N rate with slight probability of yield loss. However, the rec is often highly imprecise. Meaning that if I apply the method to 100 fields the average will be spot on, however if I look at the performance of the recommendation on a single field, I will likely be disappointed.
When it comes to nitrogen recommendations the Easy button method will use components which help ensure that the rate prescribed will maximize yield 90-95% of the time. For example, take the data presented in Figure 2. Over fifteen years of the long-term winter wheat fertility study near Lahoma, Oklahoma the average pounds of N per bushel to reach economic optimum nitrogen rate (EONR) was 1.6, however if 2.0 of N was applied per bushel yield would have been maximized 13 out of the 15 years. While 2.0 lbs. of N per bushel would have been quite accurate for maximizing yield, it would be highly imprecise as over the 15 years optimum pounds of N per bushel ranged from 0.0 to 3.2.
The trick to improving your N rate recommendation closer to a precise and accurate system is to obtain representative site-specific values for the Stanford Equation NFert = (NCrop – NSoil) / Neff.
Looking at the 15-year long-term data above the yields range from a low of 27 to a high of 88 bushels. Of those 15 years, I personally planted multiple years, usually sometime in October, and many of those years while sowing I could have guessed a range of 55-60 bushel, which just happened to be just above the 15-year average. It was not until February and March when the yield potential really started to express itself. Why, well there is a lot of weather between Oct to March, a lot of environmental positive and negative impacts on that final grain yield. This is the best timing to go out with approaches, models, or techniques to estimate yield potential for N rate recs.
While I am a big fan of soil testing, pre-plant soil samples for N are just a snap shot in time. But the While I am a big fan of soil testing, pre-plant soil samples for N are just a snapshot in time, but the nitrogen cycle Figure 3, will roar on after the soil sample is collected. Organic matter (OM) is the central component of this cycle and drives availability of NH4 and NO3 in the system. For each 1% OM in the top 6″ of the soil there is approximately 1000 lbs of organically bound N. The amount of N going into and out of OM pool is driven by C:N ratio of residues, soil temperature and soil moisture. While we very well what the mechanisms of the cycle are and can model the reactions quite well. Our inability to predict long term weather patterns is the greatest factor limiting our ability to predict future availability of NSoil.
This is where the reader should be asking “how can we get better site specific data” and I begin the discussion on why I have been promoting the of the Sensor Based Nitrogen Rate Calculator (SBNRC) and N-Rich strip method.
Lets talk about how the approach follows Stanford’s mechanistic approach to N management. First the Yield Potential component of the SBNRC which is related to NCrop. In effect researchers have built models over the past two decades that can correlate the NDVI collected from a sensor, such as the GreenSeeker, with the crops biomass and chlorophyll content. If given the number of days the crop has been growing it is possible to use the NDVI collected from the crop as a tool to predict final grain yield. The closer the wheat gets to hollow stem, or the corn gets to tassel, the better the prediction. One reason is that we have allowed more “environmental influence” to happen. Dr. Bill Raun, a founder of the SBNRC concept kept great discussion and data sets on his NUE.OKSTATE.edu website. On the “NUE Website on YP” he provides information on how yield prediction work while on the “NUE Website YP Library” he has not listed every algorithm created, and the math behind them, but also a recipe book for how anyone can create their own algorithm. While there are a lot post sensing stresses that can bring down final grain yield, the models that have been built and continually improved, do quite a good job on predicting final grain yield in-season. Resulting a much more site specific value for NCrop. The blog”Sensing the N-Rich Strip and Using the SBNRC” goes into a further discussion of using the online SBNRC.
That now leaves NSoil, which I will argue is at least as important as NCrop. As weather so greatly influences the nitrogen cycle it would be nice to have a weather station on every field paired with a 0-4 ft soil description which could be incorporated into a model. Given those might be out of reach we have found the the use of a reference strip, high N or low N, really provides an site specific estimate the of nitrogen the crop has access to. If the high N reference (N-Rich) strip is showing up that means the remainder of the field is N deficient. This may be due to losses or lack of mineralization, either way more N is needed. If the N-Rich strip is not evident then the crop is finding enough N outside of the reference strip to support its current growth. This could be that residual N or mineralization is high, or it could mean that crop growth and therefore N demand is low. Having the N check strip in each field allows for a season long evaluation. We can use NDVI to characterize how big or little of a response we have to N. We call this the Response Index (RI). An RI of 1.8 means that we could increase yield by 80% if we add adequate N, if the RI is 1.05 then we are looking at a potential increase of 5%. I have a previous blog which goes into the application of the reference strip. “Nitrogen Rich Strips, a Reminder“
Finally we combine the two, YP and RI. By predicting the yield of the area out side the N-Rich strip we can determine environmental yield potential, YP0. Basically what can the field yield if nothing is added. We multiple YP0 by the RI to get the yield potential with added N, YPN. Then its as simple as N rate = (YPN – YP0 ) x N needed per bushel. So for example if YP0 is 40 bushel RI =2, then YPN is 80 bushel. I need to fertilize the additional 40 bushels of wheat and I can use the 2.0 N per bushel can come up with a top-dress rate of 80 lbs N per acre. We are now incorporating site specific in-season NCrop and NSoil data.
And just a reminder for those of you new to my blog, I have a lot of research documenting that it is not only OK, but often best if we wait on N application in wheat and other crops. Value of In-Season N blog.
Every step we take towards the easy button is often a step towards site specific imprecision due to the use of generalized terms or models. Depending on your goals this very well could be acceptable for your operation, but with nitrogen prices as volatile as they are, should we not be considering pushing the easy button to the side, for now. Let’s add a bit of site-specific data so that we can take advantage of the N the system may be giving us, or the yield we did not expect. Let the N-Rich Strip be that first step.
Relevant Peer Review Publications.
If you have any questions please feel free to contact me @ firstname.lastname@example.org
Data used in this blog is summarized from work by
Joao Souza, under the leadership of D.B. Arnall
Lawrence Aula, under the leadership of W.R. Raun
- Wheat is highly resilient and can endure nitrogen stress for a significant period of time and fully recover.
- Delaying all nitrogen until the Feekes 5 to Feekes 7 time frame resulted in improved yields over the pre-plant 32% of the time and a loss of yield 5%. However, grain protein was improved 82% of the time with delayed nitrogen.
- It is better to delay nitrogen application to avoid conditions conducive to N loss.
Historically winter wheat producers have utilized pre-plant nitrogen (N) fertilizer application due to efficiency of time and the lower cost of the primary N source, anhydrous ammonia. However, as the growing cycle of winter wheat is approximately 9 months long with only 80% of the total N accumulation reached by flowering. Research as shown that N applied prior to planting is more likely to be lost due to leaching or denitrification. Researchers at Oklahoma State University have invested significant efforts in evaluating N management strategies. This blog will present the data from multiple trials which allowed for the comparison of nitrogen applied pre-plant versus in-season. The trials were conducted over a four-year period at multiple locations across central Oklahoma.
Delayed Nitrogen – NH4NO3
This study was started in the fall of 2016 and concluded with the 2020 wheat harvest. In all, twelve trials were established and achieved maturity. This study was designed to evaluate the recovery of winter wheat grain yield and protein after the crop was N stressed. Treatments included an untreated check, pre-plant application and ten in-season treatments. The application of in-season treatments was initiated when N deficiency was confirmed and treatments were applied in progressive order every seven growing days to the point of 63 growing days after visual deficiency (DAVD). A growing degree days is any day that the average daily temperature is at or above 40⁰ F. Ammonium nitrate (NH4NO3) was applied at a rate of 90 lbs N ac-1 for all treatments.
Nitrogen response was observed at eleven of the twelve locations, and those sites will be the focus of this review. Nitrogen applications were started ranging from Nov. 10th to Mar. 7th for 0 DAVD and, concluded with 63 DAVD occurring between mid-February and early-May. The analysis of the data evaluated the yield and protein of the in-season applications compared to both the pre-plant application and the application made at the first sign of N deficiency, 0DAVD.
Across the eleven responsive years, the pre-plant application never outperformed the 0DAVD in terms of grain yield or protein. In fact, across all location if the in-season application was made prior to the end of March, the yield and protein was equal to or better than pre-plant applications. Four out of elevens sites, yield was significantly improved with in-season applications, and protein was improved in ten out of eleven locations. For the ten site/years that had applications in March, the mid-March application of 90 lbs of N, which is about the stage of hollow stem (Feekes 6), statistically increased yield four times and protein nine times compared to the pre-plant treatment.
The studies objective was to evaluate how long the crop could be deficient and fully recover. There was no relation between when the crop became deficient and when the crop could no longer recover. Yield as maintained as long as the N was applied by late March, or just before the flag leaf is visible (Feekes 8), grain yield was the same as if applied on the first day of deficiency. However, if the N was delayed to March protein was increased six out of the eleven locations.
Delayed Nitrogen – Urea
A mirror study to the Delayed Nitrogen – NH4NO3 was established in the fall of 2018 and concluded with the 2020 harvest. This study was placed next to the NH4NO3 and treatments applied on the same days using the same rate (90 lbs N ac-1) applied as urea to evaluate efficiency of urea applications over a range of dates.
Three of the four locations produced a positive response to N fertilizer and documented similar results as the NH4NO3 project. Across these three sites in-season N was always equal to the pre-plant rate if applied before the flag leaf is visible. In addition, if the urea was applied just after hollow stem, not only was yield maintained but protein was significantly increased compared to both the pre-plant and 0DAVD treatments at all three responsive sites.
Split Rate Nitrogen – NH4NO3
This study looked at multiple rates and times of N application but for this factsheet we will focus on a small set of treatments. Performed over two years and four total sites this project looked at split application of N versus a one-time application, 45/45 split or 90 lbs of N. Application timing was 0, 30, 60, 90, 120 growing days (GDD>0), trying to have applications at planting in December, February, March and April. In three of the four sites the 90 day application produced the greatest yield and protein for both 45/45 and 90 treatments. In this study the one-time application of 90 lbs N ac-1 out yielded the 45/45 split in two of the four years and was equal the other two. The 90 day application of 90 lbs N ac-1 produced a higher protein concentration at all sites compared to the 45/45 split applied on the same date.
Nitrogen Rate by Time – Urea Source
This study evaluated four rates of N (0, 40, 80, 120 lbs N) applied at three times (30 days pre-plant, pre-plant, and Feekes 5) using urea. Feekes 5 is the growth stage prior to hollow stem when the wheats leaf sheaths are becoming strongly erect. This project was completed over two locations for two years, however of the four site/years only three statistically responded to N fertilizer. In those three responsive trials the Feekes 5 application grain yield was equal to pre-plant once, greater than pre-plant once, and less than pre-plant once. The grain protein was only statistically different between the pre-plant and Feekes 5 once, with an increase in protein with late N. The one location with yield loss can be likely attributed to N loss from urea volatilization. The urea was applied on no-till immediately after a heavy rainfall with no substantial precipitation occurring for a week after application.
This factsheet summarizes four separate research projects which can contribute data from 24 trials to evaluate the application of in-season N compared to pre-plant N, see Table 1. Of these 24 site/years we can draw conclusions from the 22 that responded to N fertilizer applications. Across these trials applying all N pre-plant resulted in the highest grain yield once, applying all N in-season near or after hollow stem resulted in an increase in grain yield above that of the pre-plant seven times. However, the delaying of N application until hollow stem resulted in a significant increase in grain protein concentration at 18 of the 22 trials.
These results are significant for the winter wheat growers of the southern Great Plains as this research documented not only the ability but the necessity to move away from pre-plant and fall N applications for winter wheat grain production. The window for N application is likely much greater than most wheat producers would have considered. This work showed that not only could N be delayed and yield not sacrificed but, when delayed; yield will be maintained and protein concentration increased.
The final conclusion is that the timing of N application should not be based upon the presence of N deficiency or calendar date. Rather the timing should be based upon the weather and enviroment during application. While many of the projects used NH4NO3 as the N source to limit the impact of N loss via volatilization, the primary source for in-season nitrogen in the region are dry urea and urea ammonium nitrate (UAN) solution. Both of these sources have well documented loss due volatilization. The location from the Nitrogen Rate by Time trial which Feekes 5 applications were statisically below the pre-plant application supports this. This data set provides signifiant evidence that the optimum application window is quite wide and allowing producers more flexiabltiy to avoid those environments which will likely lead to N losses.
Special thanks to EDC Ag Products Co LLC for suppling NH4NO3 used in the delayed N project.
For phosphorus (P) and potassium (K) fertilizer management there are three primary schools of thought when it comes to rate recommendations. The three approaches are Build-up, Maintenance/Replacement, and Sufficiency. There is a time and place for each one of the methods however the current markets are making the decision for the 2016-16 winter wheat crop a very easy one. The OSU factsheet PSS-2266 goes in-depth on each of these methods. For the rest of the blog I will use P in the conversation but in many scenarios K should/could be treated the same.
Build-up is when soil test is below a significant amount of fertilizer, about 7.5 lbs P2O5 per 1 ppm increase, is added so that soil test values increase. This method is only suggested when grain price is high and fertilizer is relatively cheap. Given the market, this is a no go. The two most commonly used methods of recommendation are Replacement and Sufficiency. In the replacement approach if the soil is at or below optimum P2O5 rate it based upon replacing what the crop will remove. The sufficiency approach uses response curves to determine the rate of P that will maximize yield. These two values are typically quite different. A good way you boil the two down is that replacement feeds the soil and sufficiency feeds the plant.
Oklahoma State Universities Soil, Water, and Forage Analytical Lab (SWFAL) provides recommendations utilizing sufficiency only while many private labs and consultants use replacement or a blended approach. Some of this is due to region. Throughout the corn belt many lease agreement contain clauses that the soil test values should not decrease otherwise the renter pays for replacement after the lease is over. For the corn belt both corn and soybean can be expected to remove 80 to 100 pounds of P per year. Conversely the Oklahoma state average wheat crop removes 17 lbs P a year. In areas where wheat yields are below 40 bushel per acre (bpa) using the sufficiency approach for P recs can increase soil test P over time.
Back to subject of this blog, consultants, agronomist, and producers need to take a good look at the way P recs are being made this year. Profitability and staying in the black is the number 1, 2, and 3 topic being discussed right now. The simple fact is there is no economic benefit to apply rate above crop need, regardless of yield level. The figures above demonstrate both the yield response to fertilizer based upon soil test. At the point of Critical level crop response / increase in yield is zero. What should also be understood is that in the replacement approach P fertilizer is still added even when soil test is in Optimum level. This also referred to as maintenance, or maintaining the current level of fertility by replacing removal. If your program is a replacement program this is not a recommendation to drop it completely. Over a period of time of high removal soil test P levels can and will be drawn down. But one year or even two years of fertilizing 100 bpa wheat based on sufficiency will not drop soil test levels. On average soils contain between 400 and 6000 pounds of total phosphorus which in the soil in three over arching forms plant available, labile, and fixed. Plant available is well plant available and fixed is non plant available. The labile form is intermediate form of P. When P is labile it can be easily converted to plant available or fixed. When a plant takes up P the system will convert labile P into available P. When we apply P fertilizer the greatest majority of was is applied makes it to the labile and fixed forms in a relatively short period of time. For more in-depth information on P in the soil you can visit the SOIL 4234 Soil Fertility course and watch recorded lectures Fall 2015 10 26-30 Link .
How to tell if your P recs have a replacement factor, not including calling your agronomist. First replacement recs are based on yield goal, so if you change your yield goal your rate will change. The other and easier way is to compare your rates to the table below. Most of the regional Land Grant Universities have very similar sufficiency recs for wheat. Another aspect of the sufficiency approach is the percent sufficiency value itself. The sufficiency can provide one more layer in the decision making process for those who are near the critical or 100% level. Response and likelihood of response to P is not equal. At the lowest levels the likelihood of response is very high and the yield increase per unit of fertilizer is the greatest. As soil test values near critical (32.5 ppm or 65 STP) the likelihood of response and amount of yield increase due to fertilizer P decreases significantly. At a STP of 10 the crop will only produce 70% of its environmental potential if P is not added while at a STP of 40 the crop will make 90% of its potential. The combination of % sufficiency and yield goal can be used to determine economic value of added P.
This year with margins tight soil testing is more important than ever before. Knowing the likelihood of response and appropriate amount of fertilizer to apply will be critical maximizing the return on fertilizer invest while maximizing the quality and amount of grain we can produce. Visit with your consultant or agronomist to discuss what the best approach is for your operation. Lets ride this market out, get the most out of every input and come out of this down cycle strong.
Feel free to contact me with any questions you may have.
With the most recent FAA UAV announcement my phone has been ringing with excited potential UAV users. Two points always comes up in the conversation. NDVI (normalized difference vegetation index) and image resolution. This blog will address the use of NDVI, resolution will come later. Before getting into the discussion, what NDVI is should be addressed. As described by Wikipedia, NDVI is a simple graphical indicator that can be used to analyze remote sensing measurements, typically but not necessarily from a space platform, and access whether the target being observed contains live green vegetation or not. NDVI is a mathematical function of the reflectance values of two wavelengths regions, near-infrared (NIR) and visable (commonly red).
The index NDVI has been tied to a great number of crop factors, the most important being biomass. Biomass being important as most things in the plant world impact biomass and biomass is related to yield. The most challenging issue with NDVI is it is highly correlated with biomass and a plants biomass is impacted by EVERYTHING!!!! Think about it, how many things can impact how a plant grows in a field.
The kicker that most do not know is that all NDVI’s values are not created equal. The source of the reflectance makes a big difference.
Measuring reflectance requires a light source, this is where the two forms of NDVI separate. Passive sensors measure reflectance using the sun (natural light) as a light source while active sensors measure the reflectance from a known light source (artificial light). The GreenSeeker is a good example of a active sensor, it emits its own light using LEDs in the sensor while satellite imagery is the classic passive sensor.
The challenge with passive remote sensing lies within the source of the light. Solar radiation and the amount of reflectance is impacted by atmospheric condition and sun angle to name a few things. That means without constant calibration, typically achieved through white plate measurements, the values are not consistent over time and space. This is the case whether the sensor is on a satellite or held held. In my research plots where I am collecting passive sensor data, so that I can measure all wavelength, I have found it necessary to collected a white plate calibration reading every 10 to 15 minutes of sensing. This is the only way I can remove the impacts of sun angle and cloud cover. When using the active sensors as long as the crop does not change the value is calibrated and repeatable.
What does this mean for those wanting to use NDVI collected from a passive sensor (satellite, plane, or UAV)? Not much if the user wants to distinguish or identify high biomass and low biomass areas. Passive NDVI is a great relative measurement for good and bad. However many who look at the measurements over time notice the values can change significantly from one day to the next. The best example I have for passive NDVI is a yield map with no legend. Even the magnitude of change between high and low is difficult to determine.
Passive NDVI in the hands of an agronomist or crop scout can be a great tool to identify zones of productivity. It becomes more complicated when decisions are made solely upon these values. One issue is this is a measure of plant biomass. It does nothing to tell us why the biomass production is different from one area to the next. That is why even with an active sensor OSU utilizes N-Rich Strips (N-Rich Strip Blog). The N-Rich Strip tells us if the difference is due to nitrogen or some other variable. We are also looking into utilizing P, K, and lime strips throughout fields. Again a good agronomist can utilize the passive NDVI data by directing sampling of the high and low biomass areas to identify the underling issues creating the differences.
OkState has been approached by many UAV companies to incorporate our nitrogen rate recommendation into their systems. This is an even greater challenge. Our sensor based nitrogen rate calculator (SBNRC blog) utilizes NDVI to predict yield based upon a model built over that last 20 years. That means to correctly work the NDVI must be calibrated and accurate to a minimum of 0.05 level (NDVI runs from 0.0 to 1.0). To date none have been able to provide a mechanism in which the NDVI could be calibrated well enough.
NDVI values collected with a passive sensor, regardless of the platform the sensor is on, has agronomic value. However its value is limited if the user is trying to make recommendations. As with any technology, to use NDVI you should have a goal in mind. It may be to identify zones or to make recommendations. Know the limitations of the technology, they all have limitations, and use the information accordingly.