Spatial Distribution And Hyperspectral Characteristics Of Heavy Metals In Cropland Soils Of Tai’an City

The research on soil pollution in Taian City is at the preliminary stage and still lacksystematic study and discussion. Main research results are as follows:（1）Seven heavy metal elements of110topsoil samples in Taian City were analyzed byfactor analysis and3main factors were selected. Combined the distribution situation ofenterprises in Taian with comparison of the3main factors, the possible origination of soilpollutants were proposed and potential ecological risk of heavy metal pollution wereevaluated. The results indicated that the heavy metal accumulation in suburb cropland wasobvious especially for Hg, Ni and Cu, with their maximum values higher than the criticalvalue of the second grade in national soil environmental quality standards. The correlationanalysis indicated that there were significant positive correlations between Zn and Ni, Zn andCu, and between Ni and Cr, while there was significant negative correlation between Zn andHg. The factor analysis results suggested that soil Zn, Cu, Ni and Cr mainly affected byindustrial and mining activities; business activity was the main factor for Pb, and agriculturalactivities and resident’s life had much more impact on the accumulation of Hg and Cd. Thepotential ecological risk assessment showed that the risks of Pb, Zn, Ni and Cr were on a lowlevel, while risks of Cd, Cu, and Hg existed with different degrees.（2）The given study has been performed in a soon-to-be urbanized region in Tai’an county,Shandong, which includes six small towns in hoping to investigate the content and the spatialdistribution of heavy metals in the crop fields around the above towns. Our study has beendone based on the geo-statistic approach by using ArcGIS with a distribution map drawn bymeans of Kriging method. The results of our investigation have shown that the actual contentof Ni、Cu and Hg in the soil is over the state soil environment quality standards level, with thevariance index of Pb and Cu being the highest, that is,50.73%for the former and43.23%forthe latter. At the same time, the variance index of Hg,32.36%, lies in the middle, while the smallest variance indexes come with Ni, Zn,Cr, Cd for their values are changing from15.72%to41.82%. The results of spatial distribution analyses show that the ratio of nugget to sill ofCu, Zn, Cd, Cr, Pb, Ni, Hg was23%,60.54%,27.88%,36.32%,63.48%,42.95%,75.61%respectively, with their range of2.0km,2.2km,15.0km,12.5km,15.6km and3.6kmrespectively. This suggests that Cu have the strongest spatial correlation, with Zn, Ni, Cd, Cr,Pb keeping the moderate spatial correlation, and Hg the feeble spatial correlation. And, finally,the spatial interpolation map reveals that the seven elements mentioned above have thedifferent spatial distribution characteristics. The high remainder of Cu, Zn, Cr, Ni content hasbeen found in the east of the region—Xintai, though the content of Hg decreases graduallyfrom the south to the north, with Cr being higher in the west and east but lower in the middle,while Pb being richer in the west and middle but fewer in the east. In addition, variousdistributions have been found influenced by the double factors: human activity for theirwell-beings and the soil itself, which suggests that we should learn to use land properly andwisely in a well-planned way for the benefit of the local people in the future.（3）In the wave band, Hyperspectral curve of soil ascended steeply, followed by tardily ascend.In the740¹080nm wave band, a high reflectivity platform was formed, and then fell down.From the1267nm beginning, the spectral curve declined gradually.（4）On the basis of hyperspetral characteristics of soil, estimation models for the Cr and Hgcontents were built after the identification of sensitive bands related to the soil Cr and Hgcontents and the construction of spectrum characteristic parameters. results showed that theoptimal Cr contents estimation model was Y=55.093+5742.9221X₃₇₄+362226.615X₁₂₆₇+399582.036X₁₂₁₃-236803.879X₂₁₄₃-80026.698X₅₁₆, which of the estimation accuracy was90.1%.The optimal Hg content estimation model was y=0.388+1.89X₂₄₉₆+145.37X₄₇₆+114.606X₄₄₆-101.581X₂₂₂₇, which of the estimation accuracy was93.0%.