Relating Crop Photosynthesis To Remotely Sensed Photochemical Reflectance Index And Sun-induced Chlorophyll Fluorescence

Gross primary production(GPP)is an important component in the global carbon cycle.It represents the photosynthetic rate of vegetation canopies,which is affected by climate change(i.e.radiation,precipitation,temperature and humidity),atmospheric CO2 concentration and atmospheric nitrogen deposition.Estimates of the spatial and temporal distributions of GPP at regional or global scales are critical to the understanding of the feedback between the carbon cycle and climate.Some models can estimate the spatial and temporal distributions of GPP at regional or global scales,but there are large differences in spatiotemporal distributions of GPP estimated from different models due to uncertainties from model structural flaws and errors from input data and parameter values.Remotely sensing signals are extensively used to estimate GPP,and these methods can be divided into three categories:(1)GPP is calculated as the product of photosynthetically active radiation(PAR),the fraction of absorbed PAR(fPAR)and light use efficiency(LUE).LUE changes with the weather conditions and some other factors.(2)The process-based models are based on the Farquhar,von Caemmerer and Berry(FvCB)photosynthesis model,which describes the photosynthesis rate mechanically at the leaf level(Farquhar et al.,1980).The leaf photosynthesis rate is dependent on the minimum of Rubisco-limited and light-limited rate.(3)GPP is directly calculated using the hyperspectral or multispectral remote sensing information.The first method is collectively named as LUE model,and this model is sensitive to LUE values.The results from different models can be significantly different.Thus,using the hyperspectral or multispectral remote sensing information to estimate GPP has received extensive attention.Photochemical reflectance index(PRI)and sun-induced chlorophyll fluorescence(SIF)can be calculated by hyperspectral remote sensing measurement.PRI is related to LUE while SIF is related to photosynthetic rate.This thesis attempts to link carbon assimilation processes to ground-based hyperspectral measurements of SIF and PRI over different species.For this purpose,the following research is conducted in this thesis:(1)canopy-level PRI and leaf-level non-photochemical quenching(NPQieaf)were measured in a maize field with five different irrigation treatments.Both PRI and NPQleaf were tested for detecting early water stress in crops;(2)Seasonal variations in leaf biochemistry(i.e.leaf chlorophyll content,leaf xanthophyll content and leaf nitrogen content)and leaf-level photosynthetic parameters(i.e.maximum carboxylation rate denoted as Vcmax25 and NPQleaf)were measured in a paddy rice field.The role of PRI was re-evaluated as an indicator of seasonally shifting pigment(chlorophyll,carotenoid and xanthophyll)contents,and photosynthetic activity for the paddy rice;(3)The red and far-red sun-induced chlorophyll fluorescence was observed from multi-angle in a wheat field.A model was used to normalize observed SIF to the hotspot direction(SIFhotspot)and to compute the canopy-level total SIF emission(SIFcanopy)in order to accurately estimate GPP of the sunlit leaf group and canopy total GPP.The major findings of this thesis are summarized as follows:(1)The quantum yield of fluorescence(0F)significantly decreased from low to moderate water stress conditions and then increased toward severe water stress conditions with soil moisture at about 20-30%of the field capacity.At the threshold of soil moisture(about 40%of the field capacity),the ratio of ?F to the quantum yield of photochemistry(?P)increased with increasing water stress,suggesting that severe drought affected ?P with higher proportions than ?F.This result indicated that the SIF would fail to indicate the photosynthetic rate when extreme drought occurs.(2)Canopy-level PRI was better than NPQleaf as an indicator of water stress at the early growing season of maize(R2 = 0.65 and p<0.05;R2 = 0.63 and p<0.05,respectively).This result encourages the use of remote sensing techniques to measure canopy-level PRI for drought-related research.However,the ability of PRI to detect water stress is confounded by many external factors(i.e.,illumination and viewing geometry).Thus,NPQieaf may be explored as a complementary parameter for detecting plant water stress.(3)Significant relationships are established between NPQleaf and Chl/Car(R2 =0.71;p<0.01)and between PRILeaf and Chl/Car(R2 = 0.58;p<0.05)at the leaf level.When water stress increased,the carotenoid contents increased while chlorophyll content remained fairly stable,leading to the decrease in the Chl/Car ratio.In the meantime,PRI also decreased,confirming that carotenoids are closely related to non-photochemical quenching in leaves,and therefore these pigments deserve close attention in water stress assessment.(4)Combining remotely sensed spectral indices(e.g.PRI)and leaf pigments(e.g.leaf chlorophyll,leaf carotenoid content,leaf xanthophyll content)can be used as good proxies for Vcmax25.Strongest correlations were found between leaf carotenoid content(Car)and Vcmax25(R2 = 0.85,p<0.001),and between leaf-level PRI*Car and Vcmax25(R2 = 0.96,p<0.001)in the growing season.These correlations provide new ways to estimate the Vcmax25 using remote sensing data because these pigment contents and spectral indices can be reliably retrieved from remote sensing data.(5)Angular normalization of multiple angle SIF measurements significantly improved the correction of SIF to GPP derived from EC measurements over a wheat canopy.The total canopy SIF(SIFcanopy)and the SIF in the hotspot direction(SIFhotspot),obtained after the angular normalization,are better correlated with GPP than the original SIF observation.The coefficients of determination(R2)increased by 0.22 ± 0.06 and 0.23 ±0.02 using SIFcanopy and SIFhotspot,respectively,after the angular normalization.In this thesis research,the relationships are investigated between photosynthesis and remotely sensed PRI and SIF for different three crop types:corn,wheat and rice.PRI and SIF measured from these crops could have different relationships with photosynthesis rate.Therefore,three experiments were designed in three different crop fields with different emphasis in order to fully explore the usefulness of remote sensing techniques for crop growth estimation.These results may have implications to crop yield monitoring at regional or global scales.