Academia.eduAcademia.edu

Figure 7 – uploaded by Zhengchao Tian

See full PDFdownloadDownload figure
Fig. 7. Comparisons of CRNS measured soil water contents (CRNS-C represents those with vegetation corrected) vs. area-average water contents determined with TDR during the 2014 and 2015 crop seasons, respectively. Two gravimetric water contents and daily precipitations are also shown.

Figure 7 Comparisons of CRNS measured soil water contents (CRNS-C represents those with vegetation corrected) vs. area-average water contents determined with TDR during the 2014 and 2015 crop seasons, respectively. Two gravimetric water contents and daily precipitations are also shown.

Related Figures (12)

Fig. 1. The locations of CRNS, TDR, and in-situ neutron probe instruments at the Lishu Experimental Station, China Agricultural University
Fig. 1. The locations of CRNS, TDR, and in-situ neutron probe instruments at the Lishu Experimental Station, China Agricultural University
Table 1
Table 1
Gravimetric water content (0m), soil organic carbon content (SOC), lattice water content (t), and bulk density (p,) of the soil samples from different layers. The samples were collected on June 25, 2014. The values in parentheses are standard deviations. after travelling a certain distance in the air. Therefore, z* also varies with the radial distance to the CRNS sensor. KOhli et al. (2015) introduced a horizontal weighting procedure to take the radial dis- tance into account. In our study, the KOhli et al. (2015) procedure was applied to estimate the depth and horizontal weighted field average 0, For details of the procedure, please refer to Kohli et al. (2015).
Gravimetric water content (0m), soil organic carbon content (SOC), lattice water content (t), and bulk density (p,) of the soil samples from different layers. The samples were collected on June 25, 2014. The values in parentheses are standard deviations. after travelling a certain distance in the air. Therefore, z* also varies with the radial distance to the CRNS sensor. KOhli et al. (2015) introduced a horizontal weighting procedure to take the radial dis- tance into account. In our study, the KOhli et al. (2015) procedure was applied to estimate the depth and horizontal weighted field average 0, For details of the procedure, please refer to Kohli et al. (2015).
Fig. 2. Water equivalent depths of biomass (BWE) and snow (SWE) vs. the average thermal/fast neutron ratios (N,) determined on the same day. Strong linear relationships were observed between the mea- sured BWE and SWE data and the corresponding daily average N, values in both 2014 and 2015 (Fig. 2). These linear relationships were applied for estimating BWE and SWE dynamics using the 12-h moving average N, data from CRNS (Fig. 3). In the 2014 crop season, the estimated BWE increased quickly with the fast growth of maize plants during the May-July period, kept a relatively high level from late July to early September (the reproductive stage), decreased gradually in mid-September to early October (the matu- rity stage), and declined rapidly in the harvest period (Fig. 3). The estimated BWE in the 2015 crop season, however, varied in a rela- tively small range because soybean field accounted for about 40% of the measurement area. The BWE of soybean field was about 70% lower than that of the maize field because soybean plant had a smaller biomass than maize plant.
Fig. 2. Water equivalent depths of biomass (BWE) and snow (SWE) vs. the average thermal/fast neutron ratios (N,) determined on the same day. Strong linear relationships were observed between the mea- sured BWE and SWE data and the corresponding daily average N, values in both 2014 and 2015 (Fig. 2). These linear relationships were applied for estimating BWE and SWE dynamics using the 12-h moving average N, data from CRNS (Fig. 3). In the 2014 crop season, the estimated BWE increased quickly with the fast growth of maize plants during the May-July period, kept a relatively high level from late July to early September (the reproductive stage), decreased gradually in mid-September to early October (the matu- rity stage), and declined rapidly in the harvest period (Fig. 3). The estimated BWE in the 2015 crop season, however, varied in a rela- tively small range because soybean field accounted for about 40% of the measurement area. The BWE of soybean field was about 70% lower than that of the maize field because soybean plant had a smaller biomass than maize plant.
Fig. 3. Daily precipitation, snow cover depth, N, method estimated dynamics of biomass water equivalent (BWE (N,)) and snow water equivalent (SWE (N,)) depths, and Baroni and Oswald (2015) method estimated biomass and snow water equivalent (WE (B&O)) depth during the study period.
Fig. 3. Daily precipitation, snow cover depth, N, method estimated dynamics of biomass water equivalent (BWE (N,)) and snow water equivalent (SWE (N,)) depths, and Baroni and Oswald (2015) method estimated biomass and snow water equivalent (WE (B&O)) depth during the study period.
Fig. 4. Fast neutron intensity, thermal neutron intensity, thermal/fast neutron ratio (N,), estimated biomass water equivalent (BWE) and snow water equivalent (SWE depths, and daily precipitation during the study period.
Fig. 4. Fast neutron intensity, thermal neutron intensity, thermal/fast neutron ratio (N,), estimated biomass water equivalent (BWE) and snow water equivalent (SWE depths, and daily precipitation during the study period.
Fig. 5. Illustration of CRNS measurements of thermal neutron intensity above the snow water layer (left) and in the biomass water layer (right). The plots on the left sides show the thermal neutron intensity gradients near the air-water interface.
Fig. 5. Illustration of CRNS measurements of thermal neutron intensity above the snow water layer (left) and in the biomass water layer (right). The plots on the left sides show the thermal neutron intensity gradients near the air-water interface.
Fig. 6. TDR measured water contents vs. thermal/fast neutron ratios (N,) during the 2014 and 2015 crop seasons, respectively. Further investigation showed that the bias of CRNS measurements (the difference between CRNS and TDR data) related closely with BWE (Fig. 8): CRNS generally overestimated THCY dosullica Uldt UNINS OVELESUITMALCU SOL Wall COLLECIIL VWWIICIT there were aboveground hydrogen pools, and AWE was related to the difference between soil water content obtained with the CRNS method and that with other independent methods (e.g., TDR and NP methods used in our study). A comparison of the AWE dynamics estimated using the Baroni and Oswald (2015) approach versus that of the N, method was presented in Fig. 3. Both datasets were obtained using a 12-h moving average of CRNS measurements. The two methods gave similar seasonal AWE trends in both crop and snow-covered seasons, but the Baroni and Oswald (2015) approach estimates displayed greater fluctuations than the N, method. The precision of the CRNS measurement increases with the number of counts due to the Poisson noises in neutron counting, which is influ- enced by the variations in solar activity, air pressure, absolute humidity, and other environmental factors (e.g., evapotranspiration and water content in plants, Bogena et al., 2013). The fluctuations of N, are relatively small because environmental factors have similar influences on both thermal and fast neutrons (Zreda et al., 2008; Desilets et al., 2010). The accuracy of the Baroni and Oswald (2015) approach is affected by the water content obtained with independent method. For example, during 3 March 2014 to 1 April 2014, the Baroni and Oswald (2015) approach provided unreason- able AWE estimates (continuous negative values), which was due to the errors in water contents obtained with NP method. Unlike the Baroni and Oswald (2015) method, soil water content had little or no effect on N, variation (Fig. 6), and the N, data were more sensi- tive to AWE than to soil water content (Fig. 2).
Fig. 6. TDR measured water contents vs. thermal/fast neutron ratios (N,) during the 2014 and 2015 crop seasons, respectively. Further investigation showed that the bias of CRNS measurements (the difference between CRNS and TDR data) related closely with BWE (Fig. 8): CRNS generally overestimated THCY dosullica Uldt UNINS OVELESUITMALCU SOL Wall COLLECIIL VWWIICIT there were aboveground hydrogen pools, and AWE was related to the difference between soil water content obtained with the CRNS method and that with other independent methods (e.g., TDR and NP methods used in our study). A comparison of the AWE dynamics estimated using the Baroni and Oswald (2015) approach versus that of the N, method was presented in Fig. 3. Both datasets were obtained using a 12-h moving average of CRNS measurements. The two methods gave similar seasonal AWE trends in both crop and snow-covered seasons, but the Baroni and Oswald (2015) approach estimates displayed greater fluctuations than the N, method. The precision of the CRNS measurement increases with the number of counts due to the Poisson noises in neutron counting, which is influ- enced by the variations in solar activity, air pressure, absolute humidity, and other environmental factors (e.g., evapotranspiration and water content in plants, Bogena et al., 2013). The fluctuations of N, are relatively small because environmental factors have similar influences on both thermal and fast neutrons (Zreda et al., 2008; Desilets et al., 2010). The accuracy of the Baroni and Oswald (2015) approach is affected by the water content obtained with independent method. For example, during 3 March 2014 to 1 April 2014, the Baroni and Oswald (2015) approach provided unreason- able AWE estimates (continuous negative values), which was due to the errors in water contents obtained with NP method. Unlike the Baroni and Oswald (2015) method, soil water content had little or no effect on N, variation (Fig. 6), and the N, data were more sensi- tive to AWE than to soil water content (Fig. 2).
Fig. 8. Differences between the CRNS and TDR measured water contents vs. biomass water equivalent (BWE) depths during the 2014 and 2015 crop seasons, respectively.
Fig. 8. Differences between the CRNS and TDR measured water contents vs. biomass water equivalent (BWE) depths during the 2014 and 2015 crop seasons, respectively.
Fig. 9. Comparisons of CRNS measured soil water contents (CRNS-C represents those with snow corrected) vs. area-average water contents determined with in-situ neutro probe (NP) during the 2013/14 and 2014/15 snow-covered seasons, respectively. The liquid water contents measured with TDR are also shown. The comparison between CRNS and NP measured soil water contents during the snow-covered seasons are presented in Fig. 9. The TDR measured area-average liquid 6, were also included in the figure. As expected, CRNS overestimated soil water contents significantly when SWE was ignored. When SWE was excluded from CRNS estimates, the RMSE between CRNS and in-situ NP data was decreased from 0.097 m? m-? to 0.039 m? m-?. Errors might occur in the area-average NP water contents because these data were derived from limited measurements at 20- and 40-cm depths. For example, during the snow-free period from 3 March 2014 to 1 April 2014, water contents obtained with the CRNS method were lower than those with the NP method. This was because after snow melting, evaporation from the top soil layer had greater influence on CRNS measurements than on NP readings obtained at 20- and 40-cm depths. We consider that the errors were small because the differences between TDR measurements and NP 0, data were
Fig. 9. Comparisons of CRNS measured soil water contents (CRNS-C represents those with snow corrected) vs. area-average water contents determined with in-situ neutro probe (NP) during the 2013/14 and 2014/15 snow-covered seasons, respectively. The liquid water contents measured with TDR are also shown. The comparison between CRNS and NP measured soil water contents during the snow-covered seasons are presented in Fig. 9. The TDR measured area-average liquid 6, were also included in the figure. As expected, CRNS overestimated soil water contents significantly when SWE was ignored. When SWE was excluded from CRNS estimates, the RMSE between CRNS and in-situ NP data was decreased from 0.097 m? m-? to 0.039 m? m-?. Errors might occur in the area-average NP water contents because these data were derived from limited measurements at 20- and 40-cm depths. For example, during the snow-free period from 3 March 2014 to 1 April 2014, water contents obtained with the CRNS method were lower than those with the NP method. This was because after snow melting, evaporation from the top soil layer had greater influence on CRNS measurements than on NP readings obtained at 20- and 40-cm depths. We consider that the errors were small because the differences between TDR measurements and NP 0, data were
Fig. 10. Differences between the CRNS and in-situ neutron probe (NP) measured water contents vs. snow water equivalent (SWE) depths during the snow-covered seasons.
Fig. 10. Differences between the CRNS and in-situ neutron probe (NP) measured water contents vs. snow water equivalent (SWE) depths during the snow-covered seasons.
Related topics:
Environmental ScienceHydrology
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved