New Criteria for Selecting Reliable Thellier-Thellier type Paleointensity Results from the 1960 Kilauea Lava Flow, Hawaii


 Thellier-Thellier type paleointensity experiments associated with partial thermal remanent magnetization (pTRM) checks have been widely used to determine paleointensity values from volcanic and archaeological media. However, previous studies further revealed that a substantial portion of paleointensity results with positive checks for historical lavas largely fails to predict the known Earth’s field intensity values. To determine the fidelity of paleointesnity values, conventional Thellier-Thellier type paleointensity experiments have been performed from the Kilauea lava flows erupted in 1960. Our results show that positive pTRM checks range from 30.34±1.39 to 53.04±1.80 μT. This strongly indicates that positive pTRM checks can’t guarantee the fidelity of paleointensity results especially when the unblocking temperatures for the newly-formed magnetic particles are higher than the treated temperature. Thus, in this study, to check thermal alteration during heating, the temperature-dependent of hysteresis parameter measured at the room-temperature for the thermally-treated samples were also measured. Our new results show that almost all biased paleointensity values correspond to Bcr /Bc >3 and CI >~10%, which indicates strong effects of domain state and thermal alteration on the fidelity of paleointensity results. Our study provides a feasible criteria to further improve the fidelity of paleointensity estimations.


Introduction
Variations in the Earth's magnetic eld intensity at different time scales bear great information on the growth of Earth's deep interiors (Macouin et al. 2004;Tarduno et al. 2007; Biggin et al. 2015) and the evolution of geodynamo (Larson and Olson, 1991;Glatzmaier et al. 1999; Olson et al. 2013). There are two different types of approaches to trace the temporal variation of geomagnetic eld intensity: relative paleointensity determination (RPI, Tauxe 1993) and absolute paleointensity determinations (Thellier 1938;Thellier and Thellier 1959). Sediments carry depositional or post-depositional remanent magnetization (DRM or pDRM), and are excellent media to records the semi-continuous RPI (Valet et  Modern absolute paleointensity determinations require multiple step heatings with systematic alteration (or consistency) checks. The Thellier protocol (Thellier 1938;Thellier and Thellier 1959) was initially proposed to compare the destruction of thermoremanent magnetization (TRM) and acquisition of laboratory-induced partial thermoremanent magnetization (pTRM) at equal temperatures. Thellier-type double heating techniques have been slightly modi ed, with each method has pros and cons (Thellier 1938;Thellier and Thellier 1959;Coe 1967;Aitken et al. 1988;Yu et al. 2004). The most commonly used technique is the so-called "Coe" protocol (Coe 1967), in which we rst heat the specimen to T i in a zeroeld to determine the NRM lost. To determine the pTRM gained, we second heat specimen to T i in an ineld conditions. Aitken et al. (1988) modi ed the Coe method (1967) by reversing the order of double heating. The IZZI protocol alternates the Aitken method in odd steps and the Coe method in even steps (Yu et al. 2004). Detailed reviews of various Thellier-type techniques were provided by Valet (2003) and Biggin (2010).
Once the paleointensity determination is carried out, the results are displayed in an Arai plot (Nagata et al. 1963) where the slope of NRM remaining versus pTRM gained is the ratio of ancient to laboratory magnetic eld intensity. Only the stable single-domain (SD) particles follow a linear Arai plots, re ecting identical spectrum of unbloking temperature (T ub ) and blocking temperature (T b ). Multi-domain (MD) and pseudo-single domain (PSD) grains tend to produce a non-linear (or sometimes zig-zagging) Arai plots The most convenient (yet the easiest) way to ensure the high delity of paleointensity determinations is to check the quality of paleointensity determination using historic rocks whose geomagnetic eld intensity is readily known (e.g., IGRF-International Geomagnetic Reference Field). To this end, the reliability of absolute paleointensity determinations was tested using historic lavas from Hawaii, US The present study was designed to incorporate the temperature dependence of magnetic hysteresis as a potential alteration checker. We aim to determine the exact mechanisms for the biased paleointensity results with positive pTRM checks. This is extremely important for paleointensity determinations for older rocks.

Samples
As the largest volcano in the world, the Kilauea is located at 19° 3' N, 204° 3' E ( Fig. 1). On January 13, 1960, eruption began along the east rift zone of Kilauea. The 1960 eruption lasted for about a month and the lava covered an area of 10 km 2 (Ritcher et al. 1970; Rowland and Walker. 1987). According to international geomagnetic reference eld (IGRF-1965) model, the geomagnetic eld intensity at the eruption site would be 36.2 µT. Paleointensity determinations have yielded results ranging from 33.55 to 53.5 µT (Tanaka and Kono 1991; Tsunakawa and Shaw 1994; Tanaka et al. 1995;and Yamamoto et al. 2003). For instance, Tanaka and Kono (1991) documented presence of two segments in an Arai plot.
They picked lower temperature segment as pTRM checks failed at higher temperature ranges. As a consequence, paleointensity determinations were overestimated (Tanaka and Kono 1991).
The rst order reversal curve (FORC) diagrams were also obtained (SF = 8), using an alternating gradient force magnetometer (MicroMag 2900). Results from the chip of A1 display closing and elongated contours, re ecting ne-grained nature (Fig. 3). It is common to observed vertical spreads on closed contours, indicating the presence of PSD (Fig. 3).

Paleointensity determination
Paleointensity determinations were accepted when they pass the following selection criteria: (1) demagnetization of NRM must be univectorial with maximum angular deviation (MAD) angles < 4° ( Kirschvink 1980); (2) pTRM checks must agree with original pTRM within 5%; (3) the ratio of maximum difference produced by a pTRM check over the length of the best-t line DRAT should be less than 10% (Selkin and Tauxe 2000); (4) the angle between the principal fractions anchored to and free from the origin DANG (Tauxe and Staudigel 2004) should be < 4°; (5) at least 6 data-points with more than 48% of extrapolated NRM fraction (Coe et al. 1978) must be included in regression analysis; (6) the degree of scatter about the best-t line normalized by the estimated slope (York 1969) should be < 5%.
Paleointensity estimates were calculated from incline of the Arai plot (Fig. 4). Paleointensity estimates were calculated from slope of Arai plot, which ratio between ancient geomagnetic eld and the applied laboratory eld. With these selection criteria, a total of 9 samples were successful paleointensity estimation ( Table 1). The magnetization removed at 560 °C (< 10%) with a few exceptions. The paleoinensity results range from 30.34±1.39 to 53.04±1.80 µT. Although consistent in range of errors (10% of IGRF model≈36.1±3.61 µT), large value differed by more than 10% from expected values. This erroneously result such as high estimated paleointensity caused by production of thermochemical chemical remanent magnetization (TCRM) during heating proccese (Yamamoto et al. 2003). µT is the temperature interval used in paleointensity estimation; n is the number of points used in paleointensity estimation; f,g, and q are NRM graction, gap factor, and quality factor of Coe et al., (1987). *is value meets the condition (B cr /B c <3, CI < 10%).

Temperature-dependent rock magnetic properties
The paleointensity results vary considerably by several factors. Non-linear feature in Arai plots can be caused either by MD effect or by thermal alteration during heating. In order to detect exact reason, sister samples were heated to the same paleointensity measuring heating steps (200, 300, 350, 400, 450, 480, 500, 520, 540, 560, 580, 600 ºC). Then after, hysteresis loops were measured at room temperature of each step to detect thermal effects (Henry et al. 2005). The room temperature normalized magnetic saturation remanence ( Figure. 5) versus temperature. The feature of Mrs T /Mrs room can be divided into two groups from room temperature to 500 °C. The rst group shows excluding low temperature relatively stable from 300 to 500 °C (Fig. 5a, e, f and g). The other group shows dome like feature (Fig. 5b, c, d, h and i). This feature mean that super-paramagnetic particles change to single domain particles or surface of pseudo single domain gained characteristic remanence magnetization at the 300 ~ 500 °C intervals. Under 300 °C points were considered to in uence of VSM. This effect is re ected in the trend of TRM-NRM trend. Basis on this, a new interval from 300 °C to 500 °C was set up and estimation of paleointensity.

Discussion
The studied of lava ows was erupted in 1960 from Hawaiian volcano. All together, 9 of 30 samples yield paleointensity results with positive pTRM checks using the IZZI protocol. The mean value of paleointensity was ~ 43.6 µT, which is higher than the expected value from the study area (~ 36.1 µT). In addition, the paleointensity results range widely between 30.34±1.39 and 53.04±1.80 µT (Fig. 6).
Previous studies have already shown that positive pTRM checks don't really guarantee the delity of paleointensity results for the Thellier-Thellier type experiments because the unblocking temperatures of the newly-formed magnetic particles could be higher than the thermally treatment temperature (Zhao et al. 2014). However, such a aw can be overcome by the room-temperature-dependent of magnetic parameters (e.g., ARM-T, SIRM-T). Upon the thermal treatment, two kinds of mineral transformations occur. The rst type of alteration can produce newly-formed strongly magnetic minerals. Despite of the distribution of the unblocking temperatures, the enhanced concentration-dependent of magnetic parameters can sensitively detect the neoformation of magnetic minerals.
The second type of thermal alterations do not produce new minerals but can oxidize the primary magnetic minerals to a less magnetic state, e.g., from magnetite to maghemite, which could signi cantly decrease the saturation magnetization (also ARM and SIRM values). Then, the rst type and second type of thermal alterations will decrease and enhance the estimated paleointensity values, respectively (Zhao et al. 2014 Our study provided a practical new criteria to determine the delity of the paleointensity especially for results with conventional positive checks. This raise a serious question for the existing paleointensity results from the geological history. We strongly suggest that domain state and thermal stability for samples should be systematically incorporated into the paleotiensity studies except for all conventional criteria.

Conclusions
We present paleointensity using a temperature dependence hysteresis parameters with modi ed Thellier experiments of IZZI protocol from the 1960 Kilauea lava. Only 9 samples of the 30 samples succeeded on the basis of the positive pTRM checks. However, a large proportion of the "successful" paleointensity results fail to predict the expected value. This strongly indicate that positive pTRM checks can't completely exclude the failure results because the conventional pTRM check can't detect the thermal alteration with the unblocking temperature of the newly-formed magnetic particles higher than the treated temperature. Therefore, we put forward that the combination of chemical index and domain state parameters should be used in the future study, especially for the older samples, to improve the delity of the paleointensity results. Declarations