Page 1 Page 2 Page 3 Page 4 Page 5 Page 6 Page 7 Page 8 Page 9 Page 10 Page 11 Page 12 Page 13 Page 14 Page 15 Page 16 Page 17 Page 18 Page 19 Page 20 Page 21 Page 22 Page 23 Page 24 Page 25 Page 26 Page 27 Page 28 Page 29 Page 30 Page 31 Page 32 Page 33 Page 34 Page 35 Page 36 Page 37 Page 38 Page 39 Page 40 Page 41 Page 42 Page 43 Page 44 Page 45 Page 46 Page 47 Page 48 Page 49 Page 50 Page 51 Page 52 Page 53 Page 54 Page 55 Page 56 Page 57 Page 58 Page 59 Page 60 Page 61 Page 62 Page 63 Page 64 Page 65 Page 66 Page 67 Page 68 Page 69 Page 70 Page 71 Page 72 Page 73 Page 74 Page 75 Page 7649 IREC Farmers' Newsletter No. 195 ­ – Rice R&D 2016 Table 1.The range of instruments used in four years’ of research to collect remotely sensed data from ‘rice variety x nitrogen’ experiments. Remote sensing Collection Bands Data collected instrument method SVC 1024 Ground Hyperspectral (330–2500 nm) 3 seasons Greenseeker Ground NDVI 2 seasons Aerial NDVI Aerial 6 bands (490–900 nm) 3 seasons micaSense Aerial 4 bands with red edge 1 season HyVista Hymap Aerial Hyperspectral (430–2450 nm) 1 season Worldview 3 Satellite 8 bands (400–1040 nm) 2 seasons l Figure 2. Measured PI nitrogen uptake (kg N/ha) versus the indices a) NDVI and b) NRENDVI collected from the micaSense RedEdge camera mounted on a drone. Remote measurement For each year of the project, a series of experiments was setup with several commercial rice varieties. Across the varieties, a range of nitrogen rates was applied to create rice plots with a large range of nitrogen uptake levels at PI. These plots were measured at PI using several remote sensing instruments and physical plant samples were also collected. The relationships between the remotely sensed data and the physically measured nitrogen uptake at PI were investigated. Multi- season correlations have been developed between the data and measured nitrogen uptake at PI for several multispectral and hyperspectral systems (Table 1). The PI nitrogen uptake prediction obtained from three years of hyperspectral data was very encouraging (see article in IREC Farmers’ Newsletter No. 194) but no hyperspectral sensors are available that are commercially viable. The four wavelengths that were most significant in the full hyperspectral prediction have been determined with the aim of testing if they alone can accurately predict PI nitrogen uptake in an aircraft-mounted filter camera. If a camera using these wavelengths can be sourced it will be used to scan plots in the coming season. We have also been evaluating the Worldview 3 satellite and the newly released micaSense RedEdge camera for their abilities to predict PI nitrogen uptake. Both of the sensors measure the red edge, which is a narrow band of reflectance (710–740 nm) that corresponds to the rapid change from low red reflectance to high near infrared (NIR) reflectance. This band is very sensitive to plant stress and provides information on the chlorophyll and nutrient status of plants. Worldview 3 Two season’s data has been collected with the very high resolution Worldview 3 satellite. The first season of the Worldview 3 satellite data showed considerable potential but when the second season’s data was analysed it showed a different relationship between PI nitrogen uptake and N2RENDVI (a normalised difference vegetation index based on the near infrared and red edge wavelengths)(Figure 1). The indice N2RENDVI calculated by ((NIR2–red edge)/(NIR2+red edge)) was shown to have the best relationship with PI nitrogen uptake. It is common for spectral data to be different across seasons and is the reason multi-season data is required when developing robust calibrations. A third season’s data will be collected in 2016–17 to determine how accurate a prediction using l Figure 1. Measured PI nitrogen uptake (kg N/ha) versus N2RENDVI for data collected from the Worldview 3 satellite for two seasons. a) NDVI NDVI NRENDVI PI N uptake (kg N/ha) PI N uptake (kg N/ha) y = -1E-05x2 + 0.004x + 0.6567 R² = 0.77463 y = 0.2125ln(x) - 0.3575 R² = 0.76807 b) NRENDVI PI N uptake (kg N/ha) R2 = 0.7363 R2 = 0.8636 N2RENDVI